jprdonnelly/MachineLearningNotebooks
1
0
Fork 0
mirror of https://github.com/Azure/MachineLearningNotebooks.git synced 2025-12-23 11:02:39 -05:00
Code Issues Projects Releases Wiki Activity

update samples from Release-139 as a part of 1.0.55 SDK release

Browse Source
This commit is contained in:
vizhur
2019-08-05 18:39:19 +00:00
parent e4d9a2b4c5
commit c0dae0c645
69 changed files with 6879 additions and 1147 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
configuration.ipynb
View File

@@ -103,7 +103,7 @@
"source": [
"import azureml.core\n",
"\n",
"print(\"This notebook was created using version 1.0.53 of the Azure ML SDK\")\n",
"print(\"This notebook was created using version 1.0.55 of the Azure ML SDK\")\n",
"print(\"You are currently using version\", azureml.core.VERSION, \"of the Azure ML SDK\")"
]
},

0
end-to-end-samples/README.md Normal file
View File

11
how-to-use-azureml/automated-machine-learning/README.md
View File

@@ -175,10 +175,19 @@ jupyter notebook
- Example of training an automated ML forecasting model on multiple time-series
- [auto-ml-classification-with-onnx.ipynb](classification-with-onnx/auto-ml-classification-with-onnx.ipynb)
- Dataset: scikit learn's [digit dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html#sklearn.datasets.load_digits)
- Dataset: scikit learn's [iris dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_iris.html)
- Simple example of using automated ML for classification with ONNX models
- Uses local compute for training
- [auto-ml-remote-amlcompute-with-onnx.ipynb](remote-amlcompute-with-onnx/auto-ml-remote-amlcompute-with-onnx.ipynb)
- Dataset: scikit learn's [iris dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_iris.html)
- Example of using automated ML for classification using remote AmlCompute for training
- Train the models with ONNX compatible config on
- Parallel execution of iterations
- Async tracking of progress
- Cancelling individual iterations or entire run
- Retrieving the ONNX models and do the inference with them
- [auto-ml-bank-marketing-subscribers-with-deployment.ipynb](bank-marketing-subscribers-with-deployment/auto-ml-bank-marketing-with-deployment.ipynb)
- Dataset: UCI's [bank marketing dataset](https://www.kaggle.com/janiobachmann/bank-marketing-dataset)
- Simple example of using automated ML for classification to predict term deposit subscriptions for a bank

4
how-to-use-azureml/automated-machine-learning/classification-bank-marketing/auto-ml-classification-bank-marketing.ipynb
View File

@@ -224,7 +224,7 @@
"outputs": [],
"source": [
"data = \"https://automlsamplenotebookdata.blob.core.windows.net/automl-sample-notebook-data/bankmarketing_train.csv\"\n",
"dflow = dprep.auto_read_file(data)\n",
"dflow = dprep.read_csv(data, infer_column_types=True)\n",
"dflow.get_profile()\n",
"X_train = dflow.drop_columns(columns=['y'])\n",
"y_train = dflow.keep_columns(columns=['y'], validate_column_exists=True)\n",
@@ -630,7 +630,7 @@
"outputs": [],
"source": [
"data = \"https://automlsamplenotebookdata.blob.core.windows.net/automl-sample-notebook-data/bankmarketing_validate.csv\"\n",
"dflow = dprep.auto_read_file(data)\n",
"dflow = dprep.read_csv(data, infer_column_types=True)\n",
"dflow.get_profile()\n",
"X_test = dflow.drop_columns(columns=['y'])\n",
"y_test = dflow.keep_columns(columns=['y'], validate_column_exists=True)\n",

2
how-to-use-azureml/automated-machine-learning/classification-credit-card-fraud/auto-ml-classification-credit-card-fraud.ipynb
View File

@@ -221,7 +221,7 @@
"outputs": [],
"source": [
"data = \"https://automlsamplenotebookdata.blob.core.windows.net/automl-sample-notebook-data/creditcard.csv\"\n",
"dflow = dprep.auto_read_file(data)\n",
"dflow = dprep.read_csv(data, infer_column_types=True)\n",
"dflow.get_profile()\n",
"X = dflow.drop_columns(columns=['Class'])\n",
"y = dflow.keep_columns(columns=['Class'], validate_column_exists=True)\n",

10
how-to-use-azureml/automated-machine-learning/classification-with-onnx/auto-ml-classification-with-onnx.ipynb
View File

@@ -29,7 +29,6 @@
"1. [Data](#Data)\n",
"1. [Train](#Train)\n",
"1. [Results](#Results)\n",
"1. [Test](#Test)\n",
"\n"
]
},
@@ -39,7 +38,7 @@
"source": [
"## Introduction\n",
"\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"In this example we use the scikit-learn's [iris dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_iris.html) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [configuration](../../../configuration.ipynb) before running this notebook.\n",
"\n",
@@ -49,7 +48,8 @@
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Configure AutoML using `AutoMLConfig`.\n",
"3. Train the model using local compute with ONNX compatible config on.\n",
"4. Explore the results and save the ONNX model."
"4. Explore the results and save the ONNX model.\n",
"5. Inference with the ONNX model."
]
},
{
@@ -156,11 +156,11 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train with enable ONNX compatible models config on\n",
"## Train\n",
"\n",
"Instantiate an `AutoMLConfig` object to specify the settings and data used to run the experiment.\n",
"\n",
"Set the parameter enable_onnx_compatible_models=True, if you also want to generate the ONNX compatible models. Please note, the forecasting task and TensorFlow models are not ONNX compatible yet.\n",
"**Note:** Set the parameter enable_onnx_compatible_models=True, if you also want to generate the ONNX compatible models. Please note, the forecasting task and TensorFlow models are not ONNX compatible yet.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",

2
how-to-use-azureml/automated-machine-learning/classification-with-whitelisting/auto-ml-classification-with-whitelisting.ipynb
View File

@@ -41,7 +41,7 @@
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [configuration](../../../configuration.ipynb) before running this notebook.\n",
"This notebooks shows how can automl can be trained on a a selected list of models,see the readme.md for the models.\n",
"This notebooks shows how can automl can be trained on a selected list of models, see the readme.md for the models.\n",
"This trains the model exclusively on tensorflow based models.\n",
"\n",
"In this notebook you will learn how to:\n",

6
how-to-use-azureml/automated-machine-learning/classification/auto-ml-classification.ipynb
View File

@@ -258,7 +258,11 @@
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"metadata": {
"tags": [
"widget-rundetails-sample"
]
},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",

2
how-to-use-azureml/automated-machine-learning/dataprep-remote-execution/auto-ml-dataprep-remote-execution.ipynb
View File

@@ -128,7 +128,7 @@
"# You can also use `read_csv` and `to_*` transformations to read (with overridable delimiter)\n",
"# and convert column types manually.\n",
"example_data = 'https://dprepdata.blob.core.windows.net/demo/crime0-random.csv'\n",
"dflow = dprep.auto_read_file(example_data).skip(1) # Remove the header row.\n",
"dflow = dprep.read_csv(example_data, infer_column_types=True)\n",
"dflow.get_profile()"
]
},

10
how-to-use-azureml/automated-machine-learning/exploring-previous-runs/auto-ml-exploring-previous-runs.ipynb
View File

@@ -197,12 +197,12 @@
"display(HTML('<h3>Iterations</h3>'))\n",
"RunDetails(ml_run).show() \n",
"\n",
"children = list(ml_run.get_children())\n",
"all_metrics = ml_run.get_metrics(recursive=True)\n",
"metricslist = {}\n",
"for run in children:\n",
" properties = run.get_properties()\n",
" metrics = {k: v for k, v in run.get_metrics().items() if isinstance(v, float)}\n",
" metricslist[int(properties['iteration'])] = metrics\n",
"for run_id, metrics in all_metrics.items():\n",
" iteration = int(run_id.split('_')[-1])\n",
" float_metrics = {k: v for k, v in metrics.items() if isinstance(v, float)}\n",
" metricslist[iteration] = float_metrics\n",
"\n",
"rundata = pd.DataFrame(metricslist).sort_index(1)\n",
"display(HTML('<h3>Metrics</h3>'))\n",

9
how-to-use-azureml/automated-machine-learning/forecasting-bike-share/auto-ml-forecasting-bike-share.ipynb
View File

@@ -345,7 +345,10 @@
"metadata": {},
"outputs": [],
"source": [
"fitted_model.named_steps['timeseriestransformer'].get_featurization_summary()"
"# Get the featurization summary as a list of JSON\n",
"featurization_summary = fitted_model.named_steps['timeseriestransformer'].get_featurization_summary()\n",
"# View the featurization summary as a pandas dataframe\n",
"pd.DataFrame.from_records(featurization_summary)"
]
},
{
@@ -522,7 +525,7 @@
"print('MAPE: %.2f' % MAPE(df_all[target_column_name], df_all['predicted']))\n",
"\n",
"# Plot outputs\n",
"%matplotlib notebook\n",
"%matplotlib inline\n",
"test_pred = plt.scatter(df_all[target_column_name], df_all['predicted'], color='b')\n",
"test_test = plt.scatter(y_test, y_test, color='g')\n",
"plt.legend((test_pred, test_test), ('prediction', 'truth'), loc='upper left', fontsize=8)\n",
@@ -564,7 +567,7 @@
"df_all_APE = df_all.assign(APE=APE(df_all[target_column_name], df_all['predicted']))\n",
"APEs = [df_all_APE[df_all['horizon_origin'] == h].APE.values for h in range(1, max_horizon + 1)]\n",
"\n",
"%matplotlib notebook\n",
"%matplotlib inline\n",
"plt.boxplot(APEs)\n",
"plt.yscale('log')\n",
"plt.xlabel('horizon')\n",

4
how-to-use-azureml/automated-machine-learning/forecasting-energy-demand/auto-ml-forecasting-energy-demand.ipynb
View File

@@ -432,7 +432,7 @@
"print('MAPE: %.2f' % MAPE(df_all[target_column_name], df_all['predicted']))\n",
"\n",
"# Plot outputs\n",
"%matplotlib notebook\n",
"%matplotlib inline\n",
"pred, = plt.plot(df_all[time_column_name], df_all['predicted'], color='b')\n",
"actual, = plt.plot(df_all[time_column_name], df_all[target_column_name], color='g')\n",
"plt.xticks(fontsize=8)\n",
@@ -543,7 +543,7 @@
"print('MAPE: %.2f' % MAPE(df_lags[target_column_name], df_lags['predicted']))\n",
"\n",
"# Plot outputs\n",
"%matplotlib notebook\n",
"%matplotlib inline\n",
"pred, = plt.plot(df_lags[time_column_name], df_lags['predicted'], color='b')\n",
"actual, = plt.plot(df_lags[time_column_name], df_lags[target_column_name], color='g')\n",
"plt.xticks(fontsize=8)\n",

2
how-to-use-azureml/automated-machine-learning/forecasting-orange-juice-sales/auto-ml-forecasting-orange-juice-sales.ipynb
View File

@@ -463,7 +463,7 @@
"# Plot outputs\n",
"import matplotlib.pyplot as plt\n",
"\n",
"%matplotlib notebook\n",
"%matplotlib inline\n",
"test_pred = plt.scatter(df_all[target_column_name], df_all['predicted'], color='b')\n",
"test_test = plt.scatter(y_test, y_test, color='g')\n",
"plt.legend((test_pred, test_test), ('prediction', 'truth'), loc='upper left', fontsize=8)\n",

5
how-to-use-azureml/automated-machine-learning/missing-data-blacklist-early-termination/auto-ml-missing-data-blacklist-early-termination.ipynb
View File

@@ -360,7 +360,10 @@
"metadata": {},
"outputs": [],
"source": [
"fitted_model.named_steps['datatransformer'].get_featurization_summary()"
"# Get the featurization summary as a list of JSON\n",
"featurization_summary = fitted_model.named_steps['datatransformer'].get_featurization_summary()\n",
"# View the featurization summary as a pandas dataframe\n",
"pd.DataFrame.from_records(featurization_summary)"
]
},
{

2
how-to-use-azureml/automated-machine-learning/regression-concrete-strength/auto-ml-regression-concrete-strength.ipynb
View File

@@ -216,7 +216,7 @@
"outputs": [],
"source": [
"data = \"https://automlsamplenotebookdata.blob.core.windows.net/automl-sample-notebook-data/compresive_strength_concrete.csv\"\n",
"dflow = dprep.auto_read_file(data)\n",
"dflow = dprep.read_csv(data, infer_column_types=True)\n",
"dflow.get_profile()\n",
"X = dflow.drop_columns(columns=['CONCRETE'])\n",
"y = dflow.keep_columns(columns=['CONCRETE'], validate_column_exists=True)\n",

2
how-to-use-azureml/automated-machine-learning/regression-hardware-performance/auto-ml-regression-hardware-performance.ipynb
View File

@@ -216,7 +216,7 @@
"outputs": [],
"source": [
"data = \"https://automlsamplenotebookdata.blob.core.windows.net/automl-sample-notebook-data/machineData.csv\"\n",
"dflow = dprep.auto_read_file(data)\n",
"dflow = dprep.read_csv(data, infer_column_types=True)\n",
"dflow.get_profile()\n",
"X = dflow.drop_columns(columns=['ERP'])\n",
"y = dflow.keep_columns(columns=['ERP'], validate_column_exists=True)\n",

554
how-to-use-azureml/automated-machine-learning/remote-amlcompute-with-onnx/auto-ml-remote-amlcompute-with-onnx.ipynb Normal file
View File

@@ -0,0 +1,554 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/automated-machine-learning/remote-amlcompute/auto-ml-remote-amlcompute.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Automated Machine Learning\n",
"_**Remote Execution using AmlCompute**_\n",
"\n",
"## Contents\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Data](#Data)\n",
"1. [Train](#Train)\n",
"1. [Results](#Results)\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"In this example we use the scikit-learn's [iris dataset](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_iris.html) to showcase how you can use AutoML for a simple classification problem.\n",
"\n",
"Make sure you have executed the [configuration](../../../configuration.ipynb) before running this notebook.\n",
"\n",
"In this notebook you would see\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Create or Attach existing AmlCompute to a workspace.\n",
"3. Configure AutoML using `AutoMLConfig`.\n",
"4. Train the model using AmlCompute with ONNX compatible config on.\n",
"5. Explore the results and save the ONNX model.\n",
"6. Inference with the ONNX model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- **Parallel** executions for iterations\n",
"- **Asynchronous** tracking of progress\n",
"- **Cancellation** of individual iterations or the entire run\n",
"- Retrieving models for any iteration or logged metric\n",
"- Specifying AutoML settings as `**kwargs`"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup\n",
"\n",
"As part of the setup you have already created an Azure ML `Workspace` object. For AutoML you will need to create an `Experiment` object, which is a named object in a `Workspace` used to run experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import logging\n",
"import os\n",
"import csv\n",
"\n",
"from matplotlib import pyplot as plt\n",
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
"from azureml.core.workspace import Workspace\n",
"from azureml.train.automl import AutoMLConfig\n",
"import azureml.dataprep as dprep"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"\n",
"# Choose a name for the run history container in the workspace.\n",
"experiment_name = 'automl-remote-amlcompute-with-onnx'\n",
"project_folder = './project'\n",
"\n",
"experiment = Experiment(ws, experiment_name)\n",
"\n",
"output = {}\n",
"output['SDK version'] = azureml.core.VERSION\n",
"output['Subscription ID'] = ws.subscription_id\n",
"output['Workspace Name'] = ws.name\n",
"output['Resource Group'] = ws.resource_group\n",
"output['Location'] = ws.location\n",
"output['Project Directory'] = project_folder\n",
"output['Experiment Name'] = experiment.name\n",
"pd.set_option('display.max_colwidth', -1)\n",
"outputDf = pd.DataFrame(data = output, index = [''])\n",
"outputDf.T"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create or Attach existing AmlCompute\n",
"You will need to create a [compute target](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#compute-target) for your AutoML run. In this tutorial, you create `AmlCompute` as your training compute resource.\n",
"\n",
"**Creation of AmlCompute takes approximately 5 minutes.** If the AmlCompute with that name is already in your workspace this code will skip the creation process.\n",
"\n",
"As with other Azure services, there are limits on certain resources (e.g. AmlCompute) associated with the Azure Machine Learning service. Please read [this article](https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-manage-quotas) on the default limits and how to request more quota."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import AmlCompute\n",
"from azureml.core.compute import ComputeTarget\n",
"\n",
"# Choose a name for your cluster.\n",
"amlcompute_cluster_name = \"cpu-cluster\"\n",
"\n",
"found = False\n",
"# Check if this compute target already exists in the workspace.\n",
"cts = ws.compute_targets\n",
"if amlcompute_cluster_name in cts and cts[amlcompute_cluster_name].type == 'AmlCompute':\n",
" found = True\n",
" print('Found existing compute target.')\n",
" compute_target = cts[amlcompute_cluster_name]\n",
"\n",
"if not found:\n",
" print('Creating a new compute target...')\n",
" provisioning_config = AmlCompute.provisioning_configuration(vm_size = \"STANDARD_D2_V2\", # for GPU, use \"STANDARD_NC6\"\n",
" #vm_priority = 'lowpriority', # optional\n",
" max_nodes = 6)\n",
"\n",
" # Create the cluster.\\n\",\n",
" compute_target = ComputeTarget.create(ws, amlcompute_cluster_name, provisioning_config)\n",
"\n",
" # Can poll for a minimum number of nodes and for a specific timeout.\n",
" # If no min_node_count is provided, it will use the scale settings for the cluster.\n",
" compute_target.wait_for_completion(show_output = True, min_node_count = None, timeout_in_minutes = 20)\n",
"\n",
" # For a more detailed view of current AmlCompute status, use get_status()."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Data\n",
"For remote executions, you need to make the data accessible from the remote compute.\n",
"This can be done by uploading the data to DataStore.\n",
"In this example, we upload scikit-learn's [load_iris](https://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_iris.html) data."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iris = datasets.load_iris()\n",
"\n",
"if not os.path.isdir('data'):\n",
" os.mkdir('data')\n",
"\n",
"if not os.path.exists(project_folder):\n",
" os.makedirs(project_folder)\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(iris.data, \n",
" iris.target, \n",
" test_size=0.2, \n",
" random_state=0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Ensure the x_train and x_test are pandas DataFrame."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Convert the X_train and X_test to pandas DataFrame and set column names,\n",
"# This is needed for initializing the input variable names of ONNX model, \n",
"# and the prediction with the ONNX model using the inference helper.\n",
"X_train = pd.DataFrame(X_train, columns=['c1', 'c2', 'c3', 'c4'])\n",
"X_test = pd.DataFrame(X_test, columns=['c1', 'c2', 'c3', 'c4'])\n",
"y_train = pd.DataFrame(y_train, columns=['label'])\n",
"\n",
"X_train.to_csv(\"data/X_train.csv\", index=False)\n",
"y_train.to_csv(\"data/y_train.csv\", index=False)\n",
"\n",
"ds = ws.get_default_datastore()\n",
"ds.upload(src_dir='./data', target_path='irisdata', overwrite=True, show_progress=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"import pkg_resources\n",
"\n",
"# create a new RunConfig object\n",
"conda_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to AmlCompute\n",
"conda_run_config.target = compute_target\n",
"conda_run_config.environment.docker.enabled = True\n",
"conda_run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"\n",
"dprep_dependency = 'azureml-dataprep==' + pkg_resources.get_distribution(\"azureml-dataprep\").version\n",
"\n",
"cd = CondaDependencies.create(pip_packages=['azureml-sdk[automl]', dprep_dependency], conda_packages=['numpy','py-xgboost<=0.80'])\n",
"conda_run_config.environment.python.conda_dependencies = cd"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Dprep reference\n",
"\n",
"Defined X and y as dprep references, which are passed to automated machine learning in the AutoMLConfig."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X = dprep.read_csv(path=ds.path('irisdata/X_train.csv'), infer_column_types=True)\n",
"y = dprep.read_csv(path=ds.path('irisdata/y_train.csv'), infer_column_types=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train\n",
"\n",
"You can specify `automl_settings` as `**kwargs` as well. Also note that you can use a `get_data()` function for local excutions too.\n",
"\n",
"**Note:** Set the parameter enable_onnx_compatible_models=True, if you also want to generate the ONNX compatible models. Please note, the forecasting task and TensorFlow models are not ONNX compatible yet.\n",
"\n",
"**Note:** When using AmlCompute, you can't pass Numpy arrays directly to the fit method.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**max_concurrent_iterations**|Maximum number of iterations that would be executed in parallel. This should be less than the number of cores on the DSVM.|\n",
"|**enable_onnx_compatible_models**|Enable the ONNX compatible models in the experiment.|"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Set the preprocess=True, currently the InferenceHelper only supports this mode."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"automl_settings = {\n",
" \"iteration_timeout_minutes\": 10,\n",
" \"iterations\": 10,\n",
" \"n_cross_validations\": 5,\n",
" \"primary_metric\": 'AUC_weighted',\n",
" \"preprocess\": True,\n",
" \"max_concurrent_iterations\": 5,\n",
" \"verbosity\": logging.INFO\n",
"}\n",
"\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
" debug_log = 'automl_errors.log',\n",
" path = project_folder,\n",
" run_configuration=conda_run_config,\n",
" X = X,\n",
" y = y,\n",
" enable_onnx_compatible_models=True, # This will generate ONNX compatible models.\n",
" **automl_settings\n",
" )"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Call the `submit` method on the experiment object and pass the run configuration. For remote runs the execution is asynchronous, so you will see the iterations get populated as they complete. You can interact with the widgets and models even when the experiment is running to retrieve the best model up to that point. Once you are satisfied with the model, you can cancel a particular iteration or the whole run.\n",
"In this example, we specify `show_output = False` to suppress console output while the run is in progress."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run = experiment.submit(automl_config, show_output = False)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Results\n",
"\n",
"#### Loading executed runs\n",
"In case you need to load a previously executed run, enable the cell below and replace the `run_id` value."
]
},
{
"cell_type": "raw",
"metadata": {},
"source": [
"remote_run = AutoMLRun(experiment = experiment, run_id = 'AutoML_5db13491-c92a-4f1d-b622-8ab8d973a058')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Widget for Monitoring Runs\n",
"\n",
"The widget will first report a \"loading\" status while running the first iteration. After completing the first iteration, an auto-updating graph and table will be shown. The widget will refresh once per minute, so you should see the graph update as child runs complete.\n",
"\n",
"You can click on a pipeline to see run properties and output logs. Logs are also available on the DSVM under `/tmp/azureml_run/{iterationid}/azureml-logs`\n",
"\n",
"**Note:** The widget displays a link at the bottom. Use this link to open a web interface to explore the individual run details."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_run"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.widgets import RunDetails\n",
"RunDetails(remote_run).show() "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Wait until the run finishes.\n",
"remote_run.wait_for_completion(show_output = True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Cancelling Runs\n",
"\n",
"You can cancel ongoing remote runs using the `cancel` and `cancel_iteration` functions."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Cancel the ongoing experiment and stop scheduling new iterations.\n",
"# remote_run.cancel()\n",
"\n",
"# Cancel iteration 1 and move onto iteration 2.\n",
"# remote_run.cancel_iteration(1)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Retrieve the Best ONNX Model\n",
"\n",
"Below we select the best pipeline from our iterations. The `get_output` method returns the best run and the fitted model. The Model includes the pipeline and any pre-processing. Overloads on `get_output` allow you to retrieve the best run and fitted model for *any* logged metric or for a particular *iteration*.\n",
"\n",
"Set the parameter return_onnx_model=True to retrieve the best ONNX model, instead of the Python model."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"best_run, onnx_mdl = remote_run.get_output(return_onnx_model=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Save the best ONNX model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.automl.core.onnx_convert import OnnxConverter\n",
"onnx_fl_path = \"./best_model.onnx\"\n",
"OnnxConverter.save_onnx_model(onnx_mdl, onnx_fl_path)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Predict with the ONNX model, using onnxruntime package"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import sys\n",
"import json\n",
"from azureml.automl.core.onnx_convert import OnnxConvertConstants\n",
"from azureml.train.automl import constants\n",
"\n",
"if sys.version_info < OnnxConvertConstants.OnnxIncompatiblePythonVersion:\n",
" python_version_compatible = True\n",
"else:\n",
" python_version_compatible = False\n",
"\n",
"try:\n",
" import onnxruntime\n",
" from azureml.automl.core.onnx_convert import OnnxInferenceHelper \n",
" onnxrt_present = True\n",
"except ImportError:\n",
" onnxrt_present = False\n",
"\n",
"def get_onnx_res(run):\n",
" res_path = 'onnx_resource.json'\n",
" run.download_file(name=constants.MODEL_RESOURCE_PATH_ONNX, output_file_path=res_path)\n",
" with open(res_path) as f:\n",
" onnx_res = json.load(f)\n",
" return onnx_res\n",
"\n",
"if onnxrt_present and python_version_compatible: \n",
" mdl_bytes = onnx_mdl.SerializeToString()\n",
" onnx_res = get_onnx_res(best_run)\n",
"\n",
" onnxrt_helper = OnnxInferenceHelper(mdl_bytes, onnx_res)\n",
" pred_onnx, pred_prob_onnx = onnxrt_helper.predict(X_test)\n",
"\n",
" print(pred_onnx)\n",
" print(pred_prob_onnx)\n",
"else:\n",
" if not python_version_compatible:\n",
" print('Please use Python version 3.6 or 3.7 to run the inference helper.') \n",
" if not onnxrt_present:\n",
" print('Please install the onnxruntime package to do the prediction with ONNX model.')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "savitam"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

9
how-to-use-azureml/automated-machine-learning/remote-amlcompute-with-onnx/auto-ml-remote-amlcompute-with-onnx.yml Normal file
View File

@@ -0,0 +1,9 @@
name: auto-ml-remote-amlcompute-with-onnx
dependencies:
- pip:
- azureml-sdk
- azureml-train-automl
- azureml-widgets
- matplotlib
- pandas_ml
- onnxruntime

4
how-to-use-azureml/automated-machine-learning/remote-amlcompute/auto-ml-remote-amlcompute.ipynb
View File

@@ -233,8 +233,8 @@
"metadata": {},
"outputs": [],
"source": [
"X = dprep.auto_read_file(path=ds.path('digitsdata/X_train.csv'))\n",
"y = dprep.auto_read_file(path=ds.path('digitsdata/y_train.csv'))"
"X = dprep.read_csv(path=ds.path('digitsdata/X_train.csv'), infer_column_types=True)\n",
"y = dprep.read_csv(path=ds.path('digitsdata/y_train.csv'), infer_column_types=True)"
]
},
{

5
how-to-use-azureml/azure-databricks/automl/automl-databricks-local-01.ipynb
View File

@@ -593,7 +593,10 @@
"metadata": {},
"outputs": [],
"source": [
"fitted_model.named_steps['datatransformer'].get_featurization_summary()"
"# Get the featurization summary as a list of JSON\n",
"featurization_summary = fitted_model.named_steps['datatransformer'].get_featurization_summary()\n",
"# View the featurization summary as a pandas dataframe\n",
"pd.DataFrame.from_records(featurization_summary)"
]
},
{

4
how-to-use-azureml/azure-databricks/databricks-as-remote-compute-target/aml-pipelines-use-databricks-as-compute-target.ipynb
View File

@@ -13,7 +13,7 @@
"metadata": {},
"source": [
"# Using Databricks as a Compute Target from Azure Machine Learning Pipeline\n",
"To use Databricks as a compute target from [Azure Machine Learning Pipeline](https://docs.microsoft.com/en-us/azure/machine-learning/service/concept-ml-pipelines), a [DatabricksStep](https://docs.microsoft.com/en-us/python/api/azureml-pipeline-steps/azureml.pipeline.steps.databricks_step.databricksstep?view=azure-ml-py) is used. This notebook demonstrates the use of DatabricksStep in Azure Machine Learning Pipeline.\n",
"To use Databricks as a compute target from [Azure Machine Learning Pipeline](https://aka.ms/pl-concept), a [DatabricksStep](https://docs.microsoft.com/en-us/python/api/azureml-pipeline-steps/azureml.pipeline.steps.databricks_step.databricksstep?view=azure-ml-py) is used. This notebook demonstrates the use of DatabricksStep in Azure Machine Learning Pipeline.\n",
"\n",
"The notebook will show:\n",
"1. Running an arbitrary Databricks notebook that the customer has in Databricks workspace\n",
@@ -675,7 +675,7 @@
"metadata": {},
"source": [
"# Next: ADLA as a Compute Target\n",
"To use ADLA as a compute target from Azure Machine Learning Pipeline, a AdlaStep is used. This [notebook](./aml-pipelines-use-adla-as-compute-target.ipynb) demonstrates the use of AdlaStep in Azure Machine Learning Pipeline."
"To use ADLA as a compute target from Azure Machine Learning Pipeline, a AdlaStep is used. This [notebook](https://aka.ms/pl-adla) demonstrates the use of AdlaStep in Azure Machine Learning Pipeline."
]
},
{

29
how-to-use-azureml/deploy-to-cloud/model-register-and-deploy.ipynb
View File

@@ -247,15 +247,38 @@
"source": [
"### Model Profiling\n",
"\n",
"you can also take advantage of profiling feature for model\n",
"You can also take advantage of the profiling feature to estimate CPU and memory requirements for models.\n",
"\n",
"```python\n",
"\n",
"profile = model.profile(ws, \"profilename\", [model], inference_config, test_sample)\n",
"profile = Model.profile(ws, \"profilename\", [model], inference_config, test_sample)\n",
"profile.wait_for_profiling(True)\n",
"profiling_results = profile.get_results()\n",
"print(profiling_results)\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Model Packaging\n",
"\n",
"If you want to build a Docker image that encapsulates your model and its dependencies, you can use the model packaging option. The output image will be pushed to your workspace's ACR.\n",
"\n",
"You must include an Environment object in your inference configuration to use `Model.package()`.\n",
"\n",
"```python\n",
"package = Model.package(ws, [model], inference_config)\n",
"package.wait_for_creation(show_output=True) # Or show_output=False to hide the Docker build logs.\n",
"package.pull()\n",
"```\n",
"\n",
"Instead of a fully-built image, you can also generate a Dockerfile and download all the assets needed to build an image on top of your Environment.\n",
"\n",
"```python\n",
"package = Model.package(ws, [model], inference_config, generate_dockerfile=True)\n",
"package.wait_for_creation(show_output=True)\n",
"package.save(\"./local_context_dir\")\n",
"```"
]
}

20
how-to-use-azureml/deploy-to-local/register-model-deploy-local-advanced.ipynb
View File

@@ -271,15 +271,10 @@
"\n",
"NOTE:\n",
"\n",
"we require docker running with linux container. If you are running Docker for Windows, you need to ensure the Linux Engine is running\n",
"The Docker image runs as a Linux container. If you are running Docker for Windows, you need to ensure the Linux Engine is running:\n",
"\n",
" powershell command to switch to linux engine\n",
" & 'C:\\Program Files\\Docker\\Docker\\DockerCli.exe' -SwitchLinuxEngine\n",
"\n",
"and c drive is shared https://docs.docker.com/docker-for-windows/#shared-drives\n",
"sometimes you have to reshare c drive as docker \n",
"\n",
"<img src=\"./dockerSharedDrive.JPG\" align=\"left\"/>"
" # PowerShell command to switch to Linux engine\n",
" & 'C:\\Program Files\\Docker\\Docker\\DockerCli.exe' -SwitchLinuxEngine"
]
},
{
@@ -295,7 +290,7 @@
"source": [
"from azureml.core.webservice import LocalWebservice\n",
"\n",
"#this is optional, if not provided we choose random port\n",
"# This is optional, if not provided Docker will choose a random unused port.\n",
"deployment_config = LocalWebservice.deploy_configuration(port=6789)\n",
"\n",
"local_service = Model.deploy(ws, \"test\", [model], inference_config, deployment_config)\n",
@@ -427,9 +422,8 @@
"local_service.reload()\n",
"print(\"--------------------------------------------------------------\")\n",
"\n",
"# after reload now if you call run this will return updated return message\n",
"\n",
"print(local_service.run(input_data=sample_input))"
"# After calling reload(), run() will return the updated message.\n",
"local_service.run(input_data=sample_input)"
]
},
{
@@ -468,7 +462,7 @@
"metadata": {
"authors": [
{
"name": "raymondl"
"name": "keriehm"
}
],
"kernelspec": {

52
how-to-use-azureml/deploy-to-local/register-model-deploy-local.ipynb
View File

@@ -106,6 +106,26 @@
" workspace=ws)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create Environment"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies\n",
"from azureml.core.environment import Environment\n",
"\n",
"environment = Environment(\"LocalDeploy\")\n",
"environment.python.conda_dependencies = CondaDependencies(\"myenv.yml\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
@@ -121,9 +141,8 @@
"source": [
"from azureml.core.model import InferenceConfig\n",
"\n",
"inference_config = InferenceConfig(runtime= \"python\", \n",
" entry_script=\"score.py\",\n",
" conda_file=\"myenv.yml\")"
"inference_config = InferenceConfig(entry_script=\"score.py\",\n",
" environment=environment)"
]
},
{
@@ -138,15 +157,10 @@
"\n",
"NOTE:\n",
"\n",
"we require docker running with linux container. If you are running Docker for Windows, you need to ensure the Linux Engine is running\n",
"The Docker image runs as a Linux container. If you are running Docker for Windows, you need to ensure the Linux Engine is running:\n",
"\n",
" powershell command to switch to linux engine\n",
" & 'C:\\Program Files\\Docker\\Docker\\DockerCli.exe' -SwitchLinuxEngine\n",
"\n",
"and c drive is shared https://docs.docker.com/docker-for-windows/#shared-drives\n",
"sometimes you have to reshare c drive as docker \n",
"\n",
"<img src=\"./dockerSharedDrive.JPG\" align=\"left\"/>"
" # PowerShell command to switch to Linux engine\n",
" & 'C:\\Program Files\\Docker\\Docker\\DockerCli.exe' -SwitchLinuxEngine"
]
},
{
@@ -157,7 +171,7 @@
"source": [
"from azureml.core.webservice import LocalWebservice\n",
"\n",
"#this is optional, if not provided we choose random port\n",
"# This is optional, if not provided Docker will choose a random unused port.\n",
"deployment_config = LocalWebservice.deploy_configuration(port=6789)\n",
"\n",
"local_service = Model.deploy(ws, \"test\", [model], inference_config, deployment_config)\n",
@@ -221,7 +235,7 @@
"\n",
"sample_input = bytes(sample_input, encoding='utf-8')\n",
"\n",
"print(local_service.run(input_data=sample_input))"
"local_service.run(input_data=sample_input)"
]
},
{
@@ -282,9 +296,8 @@
"local_service.reload()\n",
"print(\"--------------------------------------------------------------\")\n",
"\n",
"# after reload now if you call run this will return updated return message\n",
"\n",
"print(local_service.run(input_data=sample_input))"
"# After calling reload(), run() will return the updated message.\n",
"local_service.run(input_data=sample_input)"
]
},
{
@@ -296,10 +309,9 @@
"If you want to change your model(s), Conda dependencies, or deployment configuration, call `update()` to rebuild the Docker image.\n",
"\n",
"```python\n",
"\n",
"local_service.update(models=[SomeOtherModelObject],\n",
" deployment_config = local_config,\n",
" inference_config = inference_config)\n",
" inference_config=inference_config,\n",
" deployment_config=local_config)\n",
"```"
]
},
@@ -323,7 +335,7 @@
"metadata": {
"authors": [
{
"name": "raymondl"
"name": "keriehm"
}
],
"kernelspec": {

11
how-to-use-azureml/explain-model/README.md
View File

@@ -1,8 +1,11 @@
## Using explain model APIs
<a name="samples"></a>
# Explain Model SDK Sample Notebooks
Follow these sample notebooks to learn:
1. [Explain tabular data locally](explain-tabular-data-local): Basic example of explaining model trained on tabular data.
4. [Explain on remote AMLCompute](explain-on-amlcompute): Explain a model on a remote AMLCompute target.
5. [Explain tabular data with Run History](explain-tabular-data-run-history): Explain a model with Run History.
7. [Explain raw features](explain-tabular-data-raw-features): Explain the raw features of a trained model.
1. [Explain tabular data locally](tabular-data): Basic examples of explaining model trained on tabular data.
2. [Explain on remote AMLCompute](azure-integration/remote-explanation): Explain a model on a remote AMLCompute target.
3. [Explain tabular data with Run History](azure-integration/run-history): Explain a model with Run History.
4. [Operationalize model explanation](azure-integration/scoring-time): Operationalize model explanation as a web service.

645
how-to-use-azureml/explain-model/azure-integration/run-history/save-retrieve-explanations-run-history.ipynb Normal file
View File

@@ -0,0 +1,645 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/azure-integration/run-history/save-retrieve-explanations-run-history.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Save and retrieve explanations via Azure Machine Learning Run History\n",
"\n",
"_**This notebook showcases how to use the Azure Machine Learning Interpretability SDK to save and retrieve classification model explanations to/from Azure Machine Learning Run History.**_\n",
"\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Apply feature transformations\n",
" 1. Train a binary classification model\n",
" 1. Explain the model on raw features\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Upload model explanations to Azure Machine Learning Run History](#Upload)\n",
"1. [Download model explanations from Azure Machine Learning Run History](#Download)\n",
"1. [Visualize explanations](#Visualize)\n",
"1. [Next steps](#Next)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"This notebook showcases how to explain a classification model predictions locally at training time, upload explanations to the Azure Machine Learning's run history, and download previously-uploaded explanations from the Run History.\n",
"It demonstrates the API calls that you need to make to upload/download the global and local explanations and a visualization dashboard that provides an interactive way of discovering patterns in data and downloaded explanations.\n",
"\n",
"We will showcase three tabular data explainers: TabularExplainer (SHAP), MimicExplainer (global surrogate), and PFIExplainer.\n",
"\n",
"\n",
"\n",
"Problem: IBM employee attrition classification with scikit-learn (run model explainer locally and upload explanation to the Azure Machine Learning Run History)\n",
"\n",
"1. Train a SVM classification model using Scikit-learn\n",
"2. Run 'explain_model' with AML Run History, which leverages run history service to store and manage the explanation data\n",
"---\n",
"\n",
"## Setup\n",
"\n",
"You will need to have extensions enabled prior to jupyter kernel starting to see the visualization dashboard.\n",
"```\n",
"(myenv) $ jupyter nbextension install --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"(myenv) $ jupyter nbextension enable --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"```\n",
"Or\n",
"\n",
"```\n",
"(myenv) $ jupyter nbextension install azureml.contrib.explain.model.visualize --user --py\n",
"(myenv) $ jupyter nbextension enable azureml.contrib.explain.model.visualize --user --py\n",
"```\n",
"\n",
"If you are using Jupyter Labs run the following commands instead:\n",
"```\n",
"(myenv) $ jupyter labextension install @jupyter-widgets/jupyterlab-manager\n",
"(myenv) $ jupyter labextension install microsoft-mli-widget\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"\n",
"### Run model explainer locally at training time"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.pipeline import Pipeline\n",
"from sklearn.impute import SimpleImputer\n",
"from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
"from sklearn.svm import SVC\n",
"import pandas as pd\n",
"import numpy as np\n",
"\n",
"# Explainers:\n",
"# 1. SHAP Tabular Explainer\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"\n",
"# OR\n",
"\n",
"# 2. Mimic Explainer\n",
"from azureml.explain.model.mimic.mimic_explainer import MimicExplainer\n",
"# You can use one of the following four interpretable models as a global surrogate to the black box model\n",
"from azureml.explain.model.mimic.models.lightgbm_model import LGBMExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import LinearExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import SGDExplainableModel\n",
"from azureml.explain.model.mimic.models.tree_model import DecisionTreeExplainableModel\n",
"\n",
"# OR\n",
"\n",
"# 3. PFI Explainer\n",
"from azureml.explain.model.permutation.permutation_importance import PFIExplainer "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load the IBM employee attrition data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get the IBM employee attrition dataset\n",
"outdirname = 'dataset.6.21.19'\n",
"try:\n",
" from urllib import urlretrieve\n",
"except ImportError:\n",
" from urllib.request import urlretrieve\n",
"import zipfile\n",
"zipfilename = outdirname + '.zip'\n",
"urlretrieve('https://publictestdatasets.blob.core.windows.net/data/' + zipfilename, zipfilename)\n",
"with zipfile.ZipFile(zipfilename, 'r') as unzip:\n",
" unzip.extractall('.')\n",
"attritionData = pd.read_csv('./WA_Fn-UseC_-HR-Employee-Attrition.csv')\n",
"\n",
"# Dropping Employee count as all values are 1 and hence attrition is independent of this feature\n",
"attritionData = attritionData.drop(['EmployeeCount'], axis=1)\n",
"# Dropping Employee Number since it is merely an identifier\n",
"attritionData = attritionData.drop(['EmployeeNumber'], axis=1)\n",
"\n",
"attritionData = attritionData.drop(['Over18'], axis=1)\n",
"\n",
"# Since all values are 80\n",
"attritionData = attritionData.drop(['StandardHours'], axis=1)\n",
"\n",
"# Converting target variables from string to numerical values\n",
"target_map = {'Yes': 1, 'No': 0}\n",
"attritionData[\"Attrition_numerical\"] = attritionData[\"Attrition\"].apply(lambda x: target_map[x])\n",
"target = attritionData[\"Attrition_numerical\"]\n",
"\n",
"attritionXData = attritionData.drop(['Attrition_numerical', 'Attrition'], axis=1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Split data into train and test\n",
"from sklearn.model_selection import train_test_split\n",
"x_train, x_test, y_train, y_test = train_test_split(attritionXData, \n",
" target, \n",
" test_size = 0.2,\n",
" random_state=0,\n",
" stratify=target)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Creating dummy columns for each categorical feature\n",
"categorical = []\n",
"for col, value in attritionXData.iteritems():\n",
" if value.dtype == 'object':\n",
" categorical.append(col)\n",
" \n",
"# Store the numerical columns in a list numerical\n",
"numerical = attritionXData.columns.difference(categorical) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Transform raw features"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can explain raw features by either using a `sklearn.compose.ColumnTransformer` or a list of fitted transformer tuples. The cell below uses `sklearn.compose.ColumnTransformer`. In case you want to run the example with the list of fitted transformer tuples, comment the cell below and uncomment the cell that follows after. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.compose import ColumnTransformer\n",
"\n",
"# We create the preprocessing pipelines for both numeric and categorical data.\n",
"numeric_transformer = Pipeline(steps=[\n",
" ('imputer', SimpleImputer(strategy='median')),\n",
" ('scaler', StandardScaler())])\n",
"\n",
"categorical_transformer = Pipeline(steps=[\n",
" ('imputer', SimpleImputer(strategy='constant', fill_value='missing')),\n",
" ('onehot', OneHotEncoder(handle_unknown='ignore'))])\n",
"\n",
"transformations = ColumnTransformer(\n",
" transformers=[\n",
" ('num', numeric_transformer, numerical),\n",
" ('cat', categorical_transformer, categorical)])\n",
"\n",
"# Append classifier to preprocessing pipeline.\n",
"# Now we have a full prediction pipeline.\n",
"clf = Pipeline(steps=[('preprocessor', transformations),\n",
" ('classifier', SVC(kernel='linear', C = 1.0, probability=True))])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"'''\n",
"# Uncomment below if sklearn-pandas is not installed\n",
"#!pip install sklearn-pandas\n",
"from sklearn_pandas import DataFrameMapper\n",
"\n",
"# Impute, standardize the numeric features and one-hot encode the categorical features. \n",
"\n",
"\n",
"numeric_transformations = [([f], Pipeline(steps=[('imputer', SimpleImputer(strategy='median')), ('scaler', StandardScaler())])) for f in numerical]\n",
"\n",
"categorical_transformations = [([f], OneHotEncoder(handle_unknown='ignore', sparse=False)) for f in categorical]\n",
"\n",
"transformations = numeric_transformations + categorical_transformations\n",
"\n",
"# Append classifier to preprocessing pipeline.\n",
"# Now we have a full prediction pipeline.\n",
"clf = Pipeline(steps=[('preprocessor', transformations),\n",
" ('classifier', SVC(kernel='linear', C = 1.0, probability=True))]) \n",
"\n",
"\n",
"\n",
"'''"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a SVM classification model, which you want to explain"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model = clf.fit(x_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain predictions on your local machine"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# 1. Using SHAP TabularExplainer\n",
"# clf.steps[-1][1] returns the trained classification model\n",
"explainer = TabularExplainer(clf.steps[-1][1], \n",
" initialization_examples=x_train, \n",
" features=attritionXData.columns, \n",
" classes=[\"Not leaving\", \"leaving\"], \n",
" transformations=transformations)\n",
"\n",
"\n",
"\n",
"\n",
"# 2. Using MimicExplainer\n",
"# augment_data is optional and if true, oversamples the initialization examples to improve surrogate model accuracy to fit original model. Useful for high-dimensional data where the number of rows is less than the number of columns. \n",
"# max_num_of_augmentations is optional and defines max number of times we can increase the input data size.\n",
"# LGBMExplainableModel can be replaced with LinearExplainableModel, SGDExplainableModel, or DecisionTreeExplainableModel\n",
"# explainer = MimicExplainer(clf.steps[-1][1], \n",
"# x_train, \n",
"# LGBMExplainableModel, \n",
"# augment_data=True, \n",
"# max_num_of_augmentations=10, \n",
"# features=attritionXData.columns, \n",
"# classes=[\"Not leaving\", \"leaving\"], \n",
"# transformations=transformations)\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"# 3. Using PFIExplainer\n",
"\n",
"# Use the parameter \"metric\" to pass a metric name or function to evaluate the permutation. \n",
"# Note that if a metric function is provided a higher value must be better.\n",
"# Otherwise, take the negative of the function or set the parameter \"is_error_metric\" to True.\n",
"# Default metrics: \n",
"# F1 Score for binary classification, F1 Score with micro average for multiclass classification and\n",
"# Mean absolute error for regression\n",
"\n",
"# explainer = PFIExplainer(clf.steps[-1][1], \n",
"# features=x_train.columns, \n",
"# transformations=transformations,\n",
"# classes=[\"Not leaving\", \"leaving\"])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate global explanations\n",
"Explain overall model predictions (global explanation)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Passing in test dataset for evaluation examples - note it must be a representative sample of the original data\n",
"# x_train can be passed as well, but with more examples explanations will take longer although they may be more accurate\n",
"global_explanation = explainer.explain_global(x_test)\n",
"\n",
"# Note: if you used the PFIExplainer in the previous step, use the next line of code instead\n",
"# global_explanation = explainer.explain_global(x_test, true_labels=y_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Sorted SHAP values\n",
"print('ranked global importance values: {}'.format(global_explanation.get_ranked_global_values()))\n",
"# Corresponding feature names\n",
"print('ranked global importance names: {}'.format(global_explanation.get_ranked_global_names()))\n",
"# Feature ranks (based on original order of features)\n",
"print('global importance rank: {}'.format(global_explanation.global_importance_rank))\n",
"\n",
"# Note: PFIExplainer does not support per class explanations\n",
"# Per class feature names\n",
"print('ranked per class feature names: {}'.format(global_explanation.get_ranked_per_class_names()))\n",
"# Per class feature importance values\n",
"print('ranked per class feature values: {}'.format(global_explanation.get_ranked_per_class_values()))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Print out a dictionary that holds the sorted feature importance names and values\n",
"print('global importance rank: {}'.format(global_explanation.get_feature_importance_dict()))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain overall model predictions as a collection of local (instance-level) explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# feature shap values for all features and all data points in the training data\n",
"print('local importance values: {}'.format(global_explanation.local_importance_values))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate local explanations\n",
"Explain local data points (individual instances)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Note: PFIExplainer does not support local explanations\n",
"# You can pass a specific data point or a group of data points to the explain_local function\n",
"\n",
"# E.g., Explain the first data point in the test set\n",
"instance_num = 1\n",
"local_explanation = explainer.explain_local(x_test[:instance_num])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Get the prediction for the first member of the test set and explain why model made that prediction\n",
"prediction_value = clf.predict(x_test)[instance_num]\n",
"\n",
"sorted_local_importance_values = local_explanation.get_ranked_local_values()[prediction_value]\n",
"sorted_local_importance_names = local_explanation.get_ranked_local_names()[prediction_value]\n",
"\n",
"print('local importance values: {}'.format(sorted_local_importance_values))\n",
"print('local importance names: {}'.format(sorted_local_importance_names))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Upload\n",
"Upload explanations to Azure Machine Learning Run History"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import azureml.core\n",
"from azureml.core import Workspace, Experiment, Run\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"from azureml.contrib.explain.model.explanation.explanation_client import ExplanationClient\n",
"# Check core SDK version number\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ws = Workspace.from_config()\n",
"print('Workspace name: ' + ws.name, \n",
" 'Azure region: ' + ws.location, \n",
" 'Subscription id: ' + ws.subscription_id, \n",
" 'Resource group: ' + ws.resource_group, sep = '\\n')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"experiment_name = 'explain_model'\n",
"experiment = Experiment(ws, experiment_name)\n",
"run = experiment.start_logging()\n",
"client = ExplanationClient.from_run(run)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Uploading model explanation data for storage or visualization in webUX\n",
"# The explanation can then be downloaded on any compute\n",
"# Multiple explanations can be uploaded\n",
"client.upload_model_explanation(global_explanation, comment='global explanation: all features')\n",
"# Or you can only upload the explanation object with the top k feature info\n",
"#client.upload_model_explanation(global_explanation, top_k=2, comment='global explanation: Only top 2 features')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Uploading model explanation data for storage or visualization in webUX\n",
"# The explanation can then be downloaded on any compute\n",
"# Multiple explanations can be uploaded\n",
"client.upload_model_explanation(local_explanation, comment='local explanation for test point 1: all features')\n",
"\n",
"# Alterntively, you can only upload the local explanation object with the top k feature info\n",
"#client.upload_model_explanation(local_explanation, top_k=2, comment='local explanation: top 2 features')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Download\n",
"Download explanations from Azure Machine Learning Run History"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# List uploaded explanations\n",
"client.list_model_explanations()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"for explanation in client.list_model_explanations():\n",
" \n",
" if explanation['comment'] == 'local explanation for test point 1: all features':\n",
" downloaded_local_explanation = client.download_model_explanation(explanation_id=explanation['id'])\n",
" # You can pass a k value to only download the top k feature importance values\n",
" downloaded_local_explanation_top2 = client.download_model_explanation(top_k=2, explanation_id=explanation['id'])\n",
" \n",
" \n",
" elif explanation['comment'] == 'global explanation: all features':\n",
" downloaded_global_explanation = client.download_model_explanation(explanation_id=explanation['id'])\n",
" # You can pass a k value to only download the top k feature importance values\n",
" downloaded_global_explanation_top2 = client.download_model_explanation(top_k=2, explanation_id=explanation['id'])\n",
" "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Load the visualization dashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(downloaded_global_explanation, model, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
"1. [Training time: regression problem](../../tabular-data/explain-binary-classification-local.ipynb) \n",
"1. [Training time: binary classification problem](../../tabular-data/explain-binary-classification-local.ipynb)\n",
"1. [Training time: multiclass classification problem](../../tabular-data/explain-multiclass-classification-local.ipynb)\n",
"1. Explain models with engineered features:\n",
" 1. [Simple feature transformations](../../tabular-data/simple-feature-transformations-explain-local.ipynb)\n",
" 1. [Advanced feature transformations](../../tabular-data/advanced-feature-transformations-explain-local.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. Inferencing time: deploy a classification model and explainer:\n",
" 1. [Deploy a locally-trained model and explainer](../scoring-time/train-explain-model-locally-and-deploy.ipynb)\n",
" 1. [Deploy a remotely-trained model and explainer](../scoring-time/train-explain-model-on-amlcompute-and-deploy.ipynb)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

6
how-to-use-azureml/explain-model/azure-integration/run-history/save-retrieve-explanations-run-history.yml Normal file
View File

@@ -0,0 +1,6 @@
name: save-retrieve-explanations-run-history
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model

33
how-to-use-azureml/explain-model/azure-integration/scoring-time/score.py Normal file
View File

@@ -0,0 +1,33 @@
import json
import numpy as np
import pandas as pd
import os
import pickle
from sklearn.externals import joblib
from sklearn.linear_model import LogisticRegression
from azureml.core.model import Model
def init():
global original_model
global scoring_explainer
# Retrieve the path to the model file using the model name
# Assume original model is named original_prediction_model
original_model_path = Model.get_model_path('original_model')
scoring_explainer_path = Model.get_model_path('IBM_attrition_explainer')
original_model = joblib.load(original_model_path)
scoring_explainer = joblib.load(scoring_explainer_path)
def run(raw_data):
# Get predictions and explanations for each data point
data = pd.read_json(raw_data)
# Make prediction
predictions = original_model.predict(data)
# Retrieve model explanations
local_importance_values = scoring_explainer.explain(data)
# You can return any data type as long as it is JSON-serializable
return {'predictions': predictions.tolist(), 'local_importance_values': local_importance_values}

513
how-to-use-azureml/explain-model/azure-integration/scoring-time/train-explain-model-locally-and-deploy.ipynb Normal file
View File

@@ -0,0 +1,513 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/azure-integration/scoring-time/train-explain-model-locally-and-deploy.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Train and explain models locally and deploy model and scoring explainer\n",
"\n",
"\n",
"_**This notebook illustrates how to use the Azure Machine Learning Interpretability SDK to deploy a locally-trained model and its corresponding scoring explainer to Azure Container Instances (ACI) as a web service.**_\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"Problem: IBM employee attrition classification with scikit-learn (train and explain a model locally and use Azure Container Instances (ACI) for deploying your model and its corresponding scoring explainer as a web service.)\n",
"\n",
"---\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Apply feature transformations\n",
" 1. Train a binary classification model\n",
" 1. Explain the model on raw features\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Visualize explanations](#Visualize)\n",
"1. [Deploy model and scoring explainer](#Deploy)\n",
"1. [Next steps](#Next)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"\n",
"This notebook showcases how to train and explain a classification model locally, and deploy the trained model and its corresponding explainer to Azure Container Instances (ACI).\n",
"It demonstrates the API calls that you need to make to submit a run for training and explaining a model to AMLCompute, download the compute explanations remotely, and visualizing the global and local explanations via a visualization dashboard that provides an interactive way of discovering patterns in model predictions and downloaded explanations. It also demonstrates how to use Azure Machine Learning MLOps capabilities to deploy your model and its corresponding explainer.\n",
"\n",
"We will showcase one of the tabular data explainers: TabularExplainer (SHAP) and follow these steps:\n",
"1.\tDevelop a machine learning script in Python which involves the training script and the explanation script.\n",
"2.\tRun the script locally.\n",
"3.\tUse the interpretability toolkit\u00e2\u20ac\u2122s visualization dashboard to visualize predictions and their explanation. If the metrics and explanations don't indicate a desired outcome, loop back to step 1 and iterate on your scripts.\n",
"5.\tAfter a satisfactory run is found, create a scoring explainer and register the persisted model and its corresponding explainer in the model registry.\n",
"6.\tDevelop a scoring script.\n",
"7.\tCreate an image and register it in the image registry.\n",
"8.\tDeploy the image as a web service in Azure.\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup\n",
"Make sure you go through the [configuration notebook](../../../../configuration.ipynb) first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize a Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"Create An Experiment: **Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"experiment_name = 'explain_model_at_scoring_time'\n",
"experiment = Experiment(workspace=ws, name=experiment_name)\n",
"run = experiment.start_logging()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get IBM attrition data\n",
"import os\n",
"import pandas as pd\n",
"\n",
"outdirname = 'dataset.6.21.19'\n",
"try:\n",
" from urllib import urlretrieve\n",
"except ImportError:\n",
" from urllib.request import urlretrieve\n",
"import zipfile\n",
"zipfilename = outdirname + '.zip'\n",
"urlretrieve('https://publictestdatasets.blob.core.windows.net/data/' + zipfilename, zipfilename)\n",
"with zipfile.ZipFile(zipfilename, 'r') as unzip:\n",
" unzip.extractall('.')\n",
"attritionData = pd.read_csv('./WA_Fn-UseC_-HR-Employee-Attrition.csv')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.model_selection import train_test_split\n",
"from sklearn.externals import joblib\n",
"from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
"from sklearn.impute import SimpleImputer\n",
"from sklearn.pipeline import Pipeline\n",
"from sklearn.linear_model import LogisticRegression\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn_pandas import DataFrameMapper\n",
"\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"\n",
"os.makedirs('./outputs', exist_ok=True)\n",
"\n",
"# Dropping Employee count as all values are 1 and hence attrition is independent of this feature\n",
"attritionData = attritionData.drop(['EmployeeCount'], axis=1)\n",
"# Dropping Employee Number since it is merely an identifier\n",
"attritionData = attritionData.drop(['EmployeeNumber'], axis=1)\n",
"attritionData = attritionData.drop(['Over18'], axis=1)\n",
"# Since all values are 80\n",
"attritionData = attritionData.drop(['StandardHours'], axis=1)\n",
"\n",
"# Converting target variables from string to numerical values\n",
"target_map = {'Yes': 1, 'No': 0}\n",
"attritionData[\"Attrition_numerical\"] = attritionData[\"Attrition\"].apply(lambda x: target_map[x])\n",
"target = attritionData[\"Attrition_numerical\"]\n",
"\n",
"attritionXData = attritionData.drop(['Attrition_numerical', 'Attrition'], axis=1)\n",
"\n",
"# Creating dummy columns for each categorical feature\n",
"categorical = []\n",
"for col, value in attritionXData.iteritems():\n",
" if value.dtype == 'object':\n",
" categorical.append(col)\n",
"\n",
"# Store the numerical columns in a list numerical\n",
"numerical = attritionXData.columns.difference(categorical)\n",
"\n",
"numeric_transformations = [([f], Pipeline(steps=[\n",
" ('imputer', SimpleImputer(strategy='median')),\n",
" ('scaler', StandardScaler())])) for f in numerical]\n",
"\n",
"categorical_transformations = [([f], OneHotEncoder(handle_unknown='ignore', sparse=False)) for f in categorical]\n",
"\n",
"transformations = numeric_transformations + categorical_transformations\n",
"\n",
"# Append classifier to preprocessing pipeline.\n",
"# Now we have a full prediction pipeline.\n",
"clf = Pipeline(steps=[('preprocessor', DataFrameMapper(transformations)),\n",
" ('classifier', RandomForestClassifier())])\n",
"\n",
"# Split data into train and test\n",
"from sklearn.model_selection import train_test_split\n",
"x_train, x_test, y_train, y_test = train_test_split(attritionXData,\n",
" target,\n",
" test_size = 0.2,\n",
" random_state=0,\n",
" stratify=target)\n",
"\n",
"# preprocess the data and fit the classification model\n",
"clf.fit(x_train, y_train)\n",
"model = clf.steps[-1][1]\n",
"\n",
"model_file_name = 'log_reg.pkl'\n",
"\n",
"# save model in the outputs folder so it automatically get uploaded\n",
"with open(model_file_name, 'wb') as file:\n",
" joblib.dump(value=clf, filename=os.path.join('./outputs/',\n",
" model_file_name))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Explain predictions on your local machine\n",
"tabular_explainer = TabularExplainer(model, \n",
" initialization_examples=x_train, \n",
" features=attritionXData.columns, \n",
" classes=[\"Not leaving\", \"leaving\"], \n",
" transformations=transformations)\n",
"\n",
"# Explain overall model predictions (global explanation)\n",
"# Passing in test dataset for evaluation examples - note it must be a representative sample of the original data\n",
"# x_train can be passed as well, but with more examples explanations it will\n",
"# take longer although they may be more accurate\n",
"global_explanation = tabular_explainer.explain_global(x_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.explain.model.scoring.scoring_explainer import TreeScoringExplainer, save\n",
"# ScoringExplainer\n",
"scoring_explainer = TreeScoringExplainer(tabular_explainer)\n",
"# Pickle scoring explainer locally\n",
"save(scoring_explainer, exist_ok=True)\n",
"\n",
"# Register original model\n",
"run.upload_file('original_model.pkl', os.path.join('./outputs/', model_file_name))\n",
"original_model = run.register_model(model_name='original_model', model_path='original_model.pkl')\n",
"\n",
"# Register scoring explainer\n",
"run.upload_file('IBM_attrition_explainer.pkl', 'scoring_explainer.pkl')\n",
"scoring_explainer_model = run.register_model(model_name='IBM_attrition_explainer', model_path='IBM_attrition_explainer.pkl')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Visualize the explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(global_explanation, clf, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy \n",
"\n",
"Deploy Model and ScoringExplainer"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores=1, \n",
" memory_gb=1, \n",
" tags={\"data\": \"IBM_Attrition\", \n",
" \"method\" : \"local_explanation\"}, \n",
" description='Get local explanations for IBM Employee Attrition data')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"# WARNING: to install this, g++ needs to be available on the Docker image and is not by default (look at the next cell)\n",
"azureml_pip_packages = [\n",
" 'azureml-defaults', 'azureml-contrib-explain-model', 'azureml-core', 'azureml-telemetry',\n",
" 'azureml-explain-model'\n",
"]\n",
" \n",
"\n",
"# specify CondaDependencies obj\n",
"myenv = CondaDependencies.create(conda_packages=['scikit-learn', 'pandas'],\n",
" pip_packages=['sklearn-pandas', 'pyyaml'] + azureml_pip_packages,\n",
" pin_sdk_version=False)\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())\n",
"\n",
"with open(\"myenv.yml\",\"r\") as f:\n",
" print(f.read())"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile dockerfile\n",
"RUN apt-get update && apt-get install -y g++ "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.model import Model\n",
"# retrieve scoring explainer for deployment\n",
"scoring_explainer_model = Model(ws, 'IBM_attrition_explainer')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from azureml.core.image import ContainerImage\n",
"\n",
"# Use the custom scoring, docker, and conda files we created above\n",
"image_config = ContainerImage.image_configuration(execution_script=\"score.py\",\n",
" docker_file=\"dockerfile\", \n",
" runtime=\"python\", \n",
" conda_file=\"myenv.yml\")\n",
"\n",
"# Use configs and models generated above\n",
"service = Webservice.deploy_from_model(workspace=ws,\n",
" name='model-scoring',\n",
" deployment_config=aciconfig,\n",
" models=[scoring_explainer_model, original_model],\n",
" image_config=image_config)\n",
"\n",
"service.wait_for_deployment(show_output=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import requests\n",
"import json\n",
"\n",
"\n",
"# Create data to test service with\n",
"sample_data = '{\"Age\":{\"899\":49},\"BusinessTravel\":{\"899\":\"Travel_Rarely\"},\"DailyRate\":{\"899\":1098},\"Department\":{\"899\":\"Research & Development\"},\"DistanceFromHome\":{\"899\":4},\"Education\":{\"899\":2},\"EducationField\":{\"899\":\"Medical\"},\"EnvironmentSatisfaction\":{\"899\":1},\"Gender\":{\"899\":\"Male\"},\"HourlyRate\":{\"899\":85},\"JobInvolvement\":{\"899\":2},\"JobLevel\":{\"899\":5},\"JobRole\":{\"899\":\"Manager\"},\"JobSatisfaction\":{\"899\":3},\"MaritalStatus\":{\"899\":\"Married\"},\"MonthlyIncome\":{\"899\":18711},\"MonthlyRate\":{\"899\":12124},\"NumCompaniesWorked\":{\"899\":2},\"OverTime\":{\"899\":\"No\"},\"PercentSalaryHike\":{\"899\":13},\"PerformanceRating\":{\"899\":3},\"RelationshipSatisfaction\":{\"899\":3},\"StockOptionLevel\":{\"899\":1},\"TotalWorkingYears\":{\"899\":23},\"TrainingTimesLastYear\":{\"899\":2},\"WorkLifeBalance\":{\"899\":4},\"YearsAtCompany\":{\"899\":1},\"YearsInCurrentRole\":{\"899\":0},\"YearsSinceLastPromotion\":{\"899\":0},\"YearsWithCurrManager\":{\"899\":0}}'\n",
"\n",
"\n",
"\n",
"headers = {'Content-Type':'application/json'}\n",
"\n",
"# send request to service\n",
"resp = requests.post(service.scoring_uri, sample_data, headers=headers)\n",
"\n",
"print(\"POST to url\", service.scoring_uri)\n",
"# can covert back to Python objects from json string if desired\n",
"print(\"prediction:\", resp.text)\n",
"result = json.loads(resp.text)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"#plot the feature importance for the prediction\n",
"import numpy as np\n",
"import matplotlib.pyplot as plt; plt.rcdefaults()\n",
"\n",
"labels = json.loads(sample_data)\n",
"labels = labels.keys()\n",
"objects = labels\n",
"y_pos = np.arange(len(objects))\n",
"performance = result[\"local_importance_values\"][0][0]\n",
"\n",
"plt.bar(y_pos, performance, align='center', alpha=0.5)\n",
"plt.xticks(y_pos, objects)\n",
"locs, labels = plt.xticks()\n",
"plt.setp(labels, rotation=90)\n",
"plt.ylabel('Feature impact - leaving vs not leaving')\n",
"plt.title('Local feature importance for prediction')\n",
"\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
"1. [Training time: regression problem](../../tabular-data/explain-binary-classification-local.ipynb) \n",
"1. [Training time: binary classification problem](../../tabular-data/explain-binary-classification-local.ipynb)\n",
"1. [Training time: multiclass classification problem](../../tabular-data/explain-multiclass-classification-local.ipynb)\n",
"1. Explain models with engineered features:\n",
" 1. [Simple feature transformations](../../tabular-data/simple-feature-transformations-explain-local.ipynb)\n",
" 1. [Advanced feature transformations](../../tabular-data/advanced-feature-transformations-explain-local.ipynb)\n",
"1. [Save model explanations via Azure Machine Learning Run History](../run-history/save-retrieve-explanations-run-history.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. [Inferencing time: deploy a remotely-trained model and explainer](./train-explain-model-on-amlcompute-and-deploy.ipynb)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

7
how-to-use-azureml/explain-model/azure-integration/scoring-time/train-explain-model-locally-and-deploy.yml Normal file
View File

@@ -0,0 +1,7 @@
name: train-explain-model-locally-and-deploy
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model
- sklearn-pandas

548
how-to-use-azureml/explain-model/azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.ipynb Normal file
View File

@@ -0,0 +1,548 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Train and explain models remotely via Azure Machine Learning Compute and deploy model and scoring explainer\n",
"\n",
"\n",
"_**This notebook illustrates how to use the Azure Machine Learning Interpretability SDK to train and explain a classification model remotely on an Azure Machine Leanrning Compute Target (AMLCompute), and use Azure Container Instances (ACI) for deploying your model and its corresponding scoring explainer as a web service.**_\n",
"\n",
"Problem: IBM employee attrition classification with scikit-learn (train a model and run an explainer remotely via AMLCompute, and deploy model and its corresponding explainer.)\n",
"\n",
"---\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Apply feature transformations\n",
" 1. Train a binary classification model\n",
" 1. Explain the model on raw features\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Visualize results](#Visualize)\n",
"1. [Deploy model and scoring explainer](#Deploy)\n",
"1. [Next steps](#Next)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"This notebook showcases how to train and explain a classification model remotely via Azure Machine Learning Compute (AMLCompute), download the calculated explanations locally for visualization and inspection, and deploy the final model and its corresponding explainer to Azure Container Instances (ACI).\n",
"It demonstrates the API calls that you need to make to submit a run for training and explaining a model to AMLCompute, download the compute explanations remotely, and visualizing the global and local explanations via a visualization dashboard that provides an interactive way of discovering patterns in model predictions and downloaded explanations, and using Azure Machine Learning MLOps capabilities to deploy your model and its corresponding explainer.\n",
"\n",
"We will showcase one of the tabular data explainers: TabularExplainer (SHAP) and follow these steps:\n",
"1.\tDevelop a machine learning script in Python which involves the training script and the explanation script.\n",
"2.\tCreate and configure a compute target.\n",
"3.\tSubmit the scripts to the configured compute target to run in that environment. During training, the scripts can read from or write to datastore. And the records of execution (e.g., model, metrics, prediction explanations) are saved as runs in the workspace and grouped under experiments.\n",
"4.\tQuery the experiment for logged metrics and explanations from the current and past runs. Use the interpretability toolkit\u00e2\u20ac\u2122s visualization dashboard to visualize predictions and their explanation. If the metrics and explanations don't indicate a desired outcome, loop back to step 1 and iterate on your scripts.\n",
"5.\tAfter a satisfactory run is found, create a scoring explainer and register the persisted model and its corresponding explainer in the model registry.\n",
"6.\tDevelop a scoring script.\n",
"7.\tCreate an image and register it in the image registry.\n",
"8.\tDeploy the image as a web service in Azure.\n",
"\n",
"| ![azure-machine-learning-cycle](./img/azure-machine-learning-cycle.PNG) |\n",
"|:--:|"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup\n",
"Make sure you go through the [configuration notebook](../../../../configuration.ipynb) first if you haven't."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Check core SDK version number\n",
"import azureml.core\n",
"\n",
"print(\"SDK version:\", azureml.core.VERSION)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Initialize a Workspace\n",
"\n",
"Initialize a workspace object from persisted configuration"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"tags": [
"create workspace"
]
},
"outputs": [],
"source": [
"from azureml.core import Workspace\n",
"\n",
"ws = Workspace.from_config()\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep='\\n')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"\n",
"Create An Experiment: **Experiment** is a logical container in an Azure ML Workspace. It hosts run records which can include run metrics and output artifacts from your experiments."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Experiment\n",
"experiment_name = 'explainer-remote-run-on-amlcompute'\n",
"experiment = Experiment(workspace=ws, name=experiment_name)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction to AmlCompute\n",
"\n",
"Azure Machine Learning Compute is managed compute infrastructure that allows the user to easily create single to multi-node compute of the appropriate VM Family. It is created **within your workspace region** and is a resource that can be used by other users in your workspace. It autoscales by default to the max_nodes, when a job is submitted, and executes in a containerized environment packaging the dependencies as specified by the user. \n",
"\n",
"Since it is managed compute, job scheduling and cluster management are handled internally by Azure Machine Learning service. \n",
"\n",
"For more information on Azure Machine Learning Compute, please read [this article](https://docs.microsoft.com/azure/machine-learning/service/how-to-set-up-training-targets#amlcompute)\n",
"\n",
"If you are an existing BatchAI customer who is migrating to Azure Machine Learning, please read [this article](https://aka.ms/batchai-retirement)\n",
"\n",
"**Note**: As with other Azure services, there are limits on certain resources (for eg. AmlCompute quota) associated with the Azure Machine Learning service. Please read [this article](https://docs.microsoft.com/azure/machine-learning/service/how-to-manage-quotas) on the default limits and how to request more quota.\n",
"\n",
"\n",
"The training script `run_explainer.py` is already created for you. Let's have a look."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Submit an AmlCompute run in a few different ways\n",
"\n",
"First lets check which VM families are available in your region. Azure is a regional service and some specialized SKUs (especially GPUs) are only available in certain regions. Since AmlCompute is created in the region of your workspace, we will use the supported_vms () function to see if the VM family we want to use ('STANDARD_D2_V2') is supported.\n",
"\n",
"You can also pass a different region to check availability and then re-create your workspace in that region through the [configuration notebook](../../../configuration.ipynb)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import ComputeTarget, AmlCompute\n",
"\n",
"AmlCompute.supported_vmsizes(workspace=ws)\n",
"# AmlCompute.supported_vmsizes(workspace=ws, location='southcentralus')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create project directory\n",
"\n",
"Create a directory that will contain all the necessary code from your local machine that you will need access to on the remote resource. This includes the training script, and any additional files your training script depends on"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import shutil\n",
"\n",
"project_folder = './explainer-remote-run-on-amlcompute'\n",
"os.makedirs(project_folder, exist_ok=True)\n",
"shutil.copy('train_explain.py', project_folder)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Provision as a run based compute target\n",
"\n",
"You can provision AmlCompute as a compute target at run-time. In this case, the compute is auto-created for your run, scales up to max_nodes that you specify, and then **deleted automatically** after the run completes."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.runconfig import RunConfiguration\n",
"from azureml.core.conda_dependencies import CondaDependencies\n",
"from azureml.core.runconfig import DEFAULT_CPU_IMAGE\n",
"\n",
"# create a new runconfig object\n",
"run_config = RunConfiguration()\n",
"\n",
"# signal that you want to use AmlCompute to execute script.\n",
"run_config.target = \"amlcompute\"\n",
"\n",
"# AmlCompute will be created in the same region as workspace\n",
"# Set vm size for AmlCompute\n",
"run_config.amlcompute.vm_size = 'STANDARD_D2_V2'\n",
"\n",
"# enable Docker \n",
"run_config.environment.docker.enabled = True\n",
"\n",
"# set Docker base image to the default CPU-based image\n",
"run_config.environment.docker.base_image = DEFAULT_CPU_IMAGE\n",
"\n",
"# use conda_dependencies.yml to create a conda environment in the Docker image for execution\n",
"run_config.environment.python.user_managed_dependencies = False\n",
"\n",
"# auto-prepare the Docker image when used for execution (if it is not already prepared)\n",
"run_config.auto_prepare_environment = True\n",
"\n",
"azureml_pip_packages = [\n",
" 'azureml-defaults', 'azureml-contrib-explain-model', 'azureml-core', 'azureml-telemetry',\n",
" 'azureml-explain-model'\n",
"]\n",
" \n",
"\n",
"\n",
"# specify CondaDependencies obj\n",
"run_config.environment.python.conda_dependencies = CondaDependencies.create(conda_packages=['scikit-learn'],\n",
" pip_packages=['sklearn_pandas', 'pyyaml'] + azureml_pip_packages,\n",
" pin_sdk_version=False)\n",
"# Now submit a run on AmlCompute\n",
"from azureml.core.script_run_config import ScriptRunConfig\n",
"\n",
"script_run_config = ScriptRunConfig(source_directory=project_folder,\n",
" script='train_explain.py',\n",
" run_config=run_config)\n",
"\n",
"run = experiment.submit(script_run_config)\n",
"\n",
"# Show run details\n",
"run"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Note: if you need to cancel a run, you can follow [these instructions](https://aka.ms/aml-docs-cancel-run)."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%time\n",
"# Shows output of the run on stdout.\n",
"run.wait_for_completion(show_output=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Delete () is used to deprovision and delete the AmlCompute target. Useful if you want to re-use the compute name \n",
"# 'cpucluster' in this case but use a different VM family for instance.\n",
"\n",
"# cpu_cluster.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Download Model Explanation, Model, and Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# retrieve model for visualization and deployment\n",
"from azureml.core.model import Model\n",
"from sklearn.externals import joblib\n",
"original_model = Model(ws, 'original_model')\n",
"model_path = original_model.download(exist_ok=True)\n",
"original_svm_model = joblib.load(model_path)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# retrieve global explanation for visualization\n",
"from azureml.contrib.explain.model.explanation.explanation_client import ExplanationClient\n",
"\n",
"# get model explanation data\n",
"client = ExplanationClient.from_run(run)\n",
"global_explanation = client.download_model_explanation()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# retrieve x_test for visualization\n",
"from sklearn.externals import joblib\n",
"x_test_path = './x_test.pkl'\n",
"run.download_file('x_test_ibm.pkl', output_file_path=x_test_path)\n",
"x_test = joblib.load(x_test_path)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Visualize the explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(global_explanation, original_svm_model, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Deploy\n",
"Deploy Model and ScoringExplainer"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import AciWebservice\n",
"\n",
"aciconfig = AciWebservice.deploy_configuration(cpu_cores=1, \n",
" memory_gb=1, \n",
" tags={\"data\": \"IBM_Attrition\", \n",
" \"method\" : \"local_explanation\"}, \n",
" description='Get local explanations for IBM Employee Attrition data')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.conda_dependencies import CondaDependencies \n",
"\n",
"# WARNING: to install this, g++ needs to be available on the Docker image and is not by default (look at the next cell)\n",
"azureml_pip_packages = [\n",
" 'azureml-defaults', 'azureml-contrib-explain-model', 'azureml-core', 'azureml-telemetry',\n",
" 'azureml-explain-model'\n",
"]\n",
" \n",
"\n",
"# specify CondaDependencies obj\n",
"myenv = CondaDependencies.create(conda_packages=['scikit-learn', 'pandas'],\n",
" pip_packages=['sklearn-pandas', 'pyyaml'] + azureml_pip_packages,\n",
" pin_sdk_version=False)\n",
"\n",
"with open(\"myenv.yml\",\"w\") as f:\n",
" f.write(myenv.serialize_to_string())\n",
"\n",
"with open(\"myenv.yml\",\"r\") as f:\n",
" print(f.read())"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile dockerfile\n",
"RUN apt-get update && apt-get install -y g++ "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# retrieve scoring explainer for deployment\n",
"scoring_explainer_model = Model(ws, 'IBM_attrition_explainer')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.webservice import Webservice\n",
"from azureml.core.image import ContainerImage\n",
"\n",
"# Use the custom scoring, docker, and conda files we created above\n",
"image_config = ContainerImage.image_configuration(execution_script=\"score.py\",\n",
" docker_file=\"dockerfile\", \n",
" runtime=\"python\", \n",
" conda_file=\"myenv.yml\")\n",
"\n",
"# Use configs and models generated above\n",
"service = Webservice.deploy_from_model(workspace=ws,\n",
" name='model-scoring-service',\n",
" deployment_config=aciconfig,\n",
" models=[scoring_explainer_model, original_model],\n",
" image_config=image_config)\n",
"\n",
"service.wait_for_deployment(show_output=True)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import requests\n",
"\n",
"# create data to test service with\n",
"examples = x_test[:4]\n",
"input_data = examples.to_json()\n",
"\n",
"headers = {'Content-Type':'application/json'}\n",
"\n",
"# send request to service\n",
"resp = requests.post(service.scoring_uri, input_data, headers=headers)\n",
"\n",
"print(\"POST to url\", service.scoring_uri)\n",
"# can covert back to Python objects from json string if desired\n",
"print(\"prediction:\", resp.text)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"service.delete()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
"1. [Training time: regression problem](../../tabular-data/explain-binary-classification-local.ipynb) \n",
"1. [Training time: binary classification problem](../../tabular-data/explain-binary-classification-local.ipynb)\n",
"1. [Training time: multiclass classification problem](../../tabular-data/explain-multiclass-classification-local.ipynb)\n",
"1. Explain models with engineered features:\n",
" 1. [Simple feature transformations](../../tabular-data/simple-feature-transformations-explain-local.ipynb)\n",
" 1. [Advanced feature transformations](../../tabular-data/advanced-feature-transformations-explain-local.ipynb)\n",
"1. [Save model explanations via Azure Machine Learning Run History](../run-history/save-retrieve-explanations-run-history.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. [Inferencing time: deploy a locally-trained model and explainer](./train-explain-model-locally-and-deploy.ipynb)\n",
" "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

7
how-to-use-azureml/explain-model/azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.yml Normal file
View File

@@ -0,0 +1,7 @@
name: train-explain-model-on-amlcompute-and-deploy
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model
- sklearn-pandas

128
how-to-use-azureml/explain-model/azure-integration/scoring-time/train_explain.py Normal file
View File

@@ -0,0 +1,128 @@
# ---------------------------------------------------------
# Copyright (c) Microsoft Corporation. All rights reserved.
# ---------------------------------------------------------
import os
import pandas as pd
import zipfile
from sklearn.model_selection import train_test_split
from sklearn.externals import joblib
from sklearn.preprocessing import StandardScaler, OneHotEncoder
from sklearn.impute import SimpleImputer
from sklearn.pipeline import Pipeline
from sklearn.linear_model import LogisticRegression
from sklearn_pandas import DataFrameMapper
from azureml.core.run import Run
from azureml.explain.model.tabular_explainer import TabularExplainer
from azureml.contrib.explain.model.explanation.explanation_client import ExplanationClient
from azureml.explain.model.scoring.scoring_explainer import LinearScoringExplainer, save
OUTPUT_DIR = './outputs/'
os.makedirs(OUTPUT_DIR, exist_ok=True)
# get the IBM employee attrition dataset
outdirname = 'dataset.6.21.19'
try:
from urllib import urlretrieve
except ImportError:
from urllib.request import urlretrieve
zipfilename = outdirname + '.zip'
urlretrieve('https://publictestdatasets.blob.core.windows.net/data/' + zipfilename, zipfilename)
with zipfile.ZipFile(zipfilename, 'r') as unzip:
unzip.extractall('.')
attritionData = pd.read_csv('./WA_Fn-UseC_-HR-Employee-Attrition.csv')
# dropping Employee count as all values are 1 and hence attrition is independent of this feature
attritionData = attritionData.drop(['EmployeeCount'], axis=1)
# dropping Employee Number since it is merely an identifier
attritionData = attritionData.drop(['EmployeeNumber'], axis=1)
attritionData = attritionData.drop(['Over18'], axis=1)
# since all values are 80
attritionData = attritionData.drop(['StandardHours'], axis=1)
# converting target variables from string to numerical values
target_map = {'Yes': 1, 'No': 0}
attritionData["Attrition_numerical"] = attritionData["Attrition"].apply(lambda x: target_map[x])
target = attritionData["Attrition_numerical"]
attritionXData = attritionData.drop(['Attrition_numerical', 'Attrition'], axis=1)
# creating dummy columns for each categorical feature
categorical = []
for col, value in attritionXData.iteritems():
if value.dtype == 'object':
categorical.append(col)
# store the numerical columns
numerical = attritionXData.columns.difference(categorical)
numeric_transformations = [([f], Pipeline(steps=[
('imputer', SimpleImputer(strategy='median')),
('scaler', StandardScaler())])) for f in numerical]
categorical_transformations = [([f], OneHotEncoder(handle_unknown='ignore', sparse=False)) for f in categorical]
transformations = numeric_transformations + categorical_transformations
# append classifier to preprocessing pipeline
clf = Pipeline(steps=[('preprocessor', DataFrameMapper(transformations)),
('classifier', LogisticRegression(solver='lbfgs'))])
# get the run this was submitted from to interact with run history
run = Run.get_context()
# create an explanation client to store the explanation (contrib API)
client = ExplanationClient.from_run(run)
# Split data into train and test
x_train, x_test, y_train, y_test = train_test_split(attritionXData,
target,
test_size=0.2,
random_state=0,
stratify=target)
# write x_test out as a pickle file for later visualization
x_test_pkl = 'x_test.pkl'
with open(x_test_pkl, 'wb') as file:
joblib.dump(value=x_test, filename=os.path.join(OUTPUT_DIR, x_test_pkl))
run.upload_file('x_test_ibm.pkl', os.path.join(OUTPUT_DIR, x_test_pkl))
# preprocess the data and fit the classification model
clf.fit(x_train, y_train)
model = clf.steps[-1][1]
# save model for use outside the script
model_file_name = 'log_reg.pkl'
with open(model_file_name, 'wb') as file:
joblib.dump(value=clf, filename=os.path.join(OUTPUT_DIR, model_file_name))
# register the model with the model management service for later use
run.upload_file('original_model.pkl', os.path.join(OUTPUT_DIR, model_file_name))
original_model = run.register_model(model_name='original_model', model_path='original_model.pkl')
# create an explainer to validate or debug the model
tabular_explainer = TabularExplainer(model,
initialization_examples=x_train,
features=attritionXData.columns,
classes=["Not leaving", "leaving"],
transformations=transformations)
# explain overall model predictions (global explanation)
# passing in test dataset for evaluation examples - note it must be a representative sample of the original data
# more data (e.g. x_train) will likely lead to higher accuracy, but at a time cost
global_explanation = tabular_explainer.explain_global(x_test)
# uploading model explanation data for storage or visualization
comment = 'Global explanation on classification model trained on IBM employee attrition dataset'
client.upload_model_explanation(global_explanation, comment=comment)
# also create a lightweight explainer for scoring time
scoring_explainer = LinearScoringExplainer(tabular_explainer)
# pickle scoring explainer locally
save(scoring_explainer, directory=OUTPUT_DIR, exist_ok=True)
# register scoring explainer
run.upload_file('IBM_attrition_explainer.pkl', os.path.join(OUTPUT_DIR, 'scoring_explainer.pkl'))
scoring_explainer_model = run.register_model(model_name='IBM_attrition_explainer',
model_path='IBM_attrition_explainer.pkl')

523
how-to-use-azureml/explain-model/tabular-data/advanced-feature-transformations-explain-local.ipynb Normal file
View File

@@ -0,0 +1,523 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/tabular-data/advanced-feature-transformations-explain-local.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Explain binary classification model predictions with raw feature transformations\n",
"_**This notebook showcases how to use the Azure Machine Learning Interpretability SDK to explain and visualize a binary classification model that uses advanced many to one or many to many feature transformations.**_\n",
"\n",
"\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Apply feature transformations\n",
" 1. Train a binary classification model\n",
" 1. Explain the model on raw features\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Visualize results](#Visualize)\n",
"1. [Next steps](#Next)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"This notebook illustrates creating explanations for a binary classification model, Titanic passenger data classification, that uses many to one and many to many feature transformations from raw data to engineered features. For the many to one transformation, we sum 2 features `age` and `fare`. For many to many transformations two features are computed: one that is product of `age` and `fare` and another that is square of this product. Our tabular data explainer is then used to get the explanation object with the flag `allow_all_transformations` passed. The object is then used to get raw feature importances.\n",
"\n",
"\n",
"We will showcase raw feature transformations with three tabular data explainers: TabularExplainer (SHAP), MimicExplainer (global surrogate), and PFIExplainer.\n",
"\n",
"| ![Interpretability Toolkit Architecture](./img/interpretability-architecture.PNG) |\n",
"|:--:|\n",
"| *Interpretability Toolkit Architecture* |\n",
"\n",
"Problem: Titanic passenger data classification with scikit-learn (run model explainer locally)\n",
"\n",
"1. Transform raw features to engineered features\n",
"2. Train a Logistic Regression model using Scikit-learn\n",
"3. Run 'explain_model' globally and locally with full dataset in local mode, which doesn't contact any Azure services.\n",
"4. Visualize the global and local explanations with the visualization dashboard.\n",
"---\n",
"\n",
"## Setup\n",
"\n",
"You will need to have extensions enabled prior to jupyter kernel starting to see the visualization dashboard.\n",
"```\n",
"(myenv) $ jupyter nbextension install --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"(myenv) $ jupyter nbextension enable --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"```\n",
"Or\n",
"\n",
"```\n",
"(myenv) $ jupyter nbextension install azureml.contrib.explain.model.visualize --user --py\n",
"(myenv) $ jupyter nbextension enable azureml.contrib.explain.model.visualize --user --py\n",
"```\n",
"\n",
"If you are using Jupyter Labs run the following commands instead:\n",
"```\n",
"(myenv) $ jupyter labextension install @jupyter-widgets/jupyterlab-manager\n",
"(myenv) $ jupyter labextension install microsoft-mli-widget\n",
"```\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"\n",
"### Run model explainer locally at training time"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.pipeline import Pipeline\n",
"from sklearn.impute import SimpleImputer\n",
"from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
"from sklearn.linear_model import LogisticRegression\n",
"import pandas as pd\n",
"import numpy as np\n",
"\n",
"# Explainers:\n",
"# 1. SHAP Tabular Explainer\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"\n",
"# OR\n",
"\n",
"# 2. Mimic Explainer\n",
"from azureml.explain.model.mimic.mimic_explainer import MimicExplainer\n",
"# You can use one of the following four interpretable models as a global surrogate to the black box model\n",
"from azureml.explain.model.mimic.models.lightgbm_model import LGBMExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import LinearExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import SGDExplainableModel\n",
"from azureml.explain.model.mimic.models.tree_model import DecisionTreeExplainableModel\n",
"\n",
"# OR\n",
"\n",
"# 3. PFI Explainer\n",
"from azureml.explain.model.permutation.permutation_importance import PFIExplainer "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load the Titanic passenger data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"titanic_url = ('https://raw.githubusercontent.com/amueller/'\n",
" 'scipy-2017-sklearn/091d371/notebooks/datasets/titanic3.csv')\n",
"data = pd.read_csv(titanic_url)\n",
"# fill missing values\n",
"data = data.fillna(method=\"ffill\")\n",
"data = data.fillna(method=\"bfill\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Similar to example [here](https://scikit-learn.org/stable/auto_examples/compose/plot_column_transformer_mixed_types.html#sphx-glr-auto-examples-compose-plot-column-transformer-mixed-types-py), use a subset of columns"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.model_selection import train_test_split\n",
"\n",
"numeric_features = ['age', 'fare']\n",
"categorical_features = ['embarked', 'sex', 'pclass']\n",
"\n",
"y = data['survived'].values\n",
"X = data[categorical_features + numeric_features]\n",
"\n",
"# Split data into train and test\n",
"x_train, x_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Transform raw features"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can explain raw features by either using a `sklearn.compose.ColumnTransformer` or a list of fitted transformer tuples. The cell below uses `sklearn.compose.ColumnTransformer`. In case you want to run the example with the list of fitted transformer tuples, comment the cell below and uncomment the cell that follows after. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# We add many to one and many to many transformations for illustration purposes.\n",
"# The support for raw feature explanations with many to one and many to many transformations are only supported \n",
"# When allow_all_transformations is set to True on explainer creation\n",
"from sklearn.preprocessing import FunctionTransformer\n",
"many_to_one_transformer = FunctionTransformer(lambda x: x.sum(axis=1).reshape(-1, 1))\n",
"many_to_many_transformer = FunctionTransformer(lambda x: np.hstack(\n",
" (np.prod(x, axis=1).reshape(-1, 1), (np.prod(x, axis=1)**2).reshape(-1, 1))\n",
"))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.compose import ColumnTransformer\n",
"\n",
"transformations = ColumnTransformer([\n",
" (\"age_fare_1\", Pipeline(steps=[\n",
" ('imputer', SimpleImputer(strategy='median')),\n",
" ('scaler', StandardScaler())\n",
" ]), [\"age\", \"fare\"]),\n",
" (\"age_fare_2\", many_to_one_transformer, [\"age\", \"fare\"]),\n",
" (\"age_fare_3\", many_to_many_transformer, [\"age\", \"fare\"]),\n",
" (\"embarked\", Pipeline(steps=[\n",
" (\"imputer\", SimpleImputer(strategy='constant', fill_value='missing')), \n",
" (\"encoder\", OneHotEncoder(sparse=False))]), [\"embarked\"]),\n",
" (\"sex_pclass\", OneHotEncoder(sparse=False), [\"sex\", \"pclass\"]) \n",
"])\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"'''\n",
"# Uncomment below if sklearn-pandas is not installed\n",
"#!pip install sklearn-pandas\n",
"from sklearn_pandas import DataFrameMapper\n",
"\n",
"# Impute, standardize the numeric features and one-hot encode the categorical features. \n",
"\n",
"transformations = [\n",
" ([\"age\", \"fare\"], Pipeline(steps=[\n",
" ('imputer', SimpleImputer(strategy='median')),\n",
" ('scaler', StandardScaler())\n",
" ])),\n",
" ([\"age\", \"fare\"], many_to_one_transformer),\n",
" ([\"age\", \"fare\"], many_to_many_transformer),\n",
" ([\"embarked\"], Pipeline(steps=[\n",
" (\"imputer\", SimpleImputer(strategy='constant', fill_value='missing')), \n",
" (\"encoder\", OneHotEncoder(sparse=False))])),\n",
" ([\"sex\", \"pclass\"], OneHotEncoder(sparse=False)) \n",
"]\n",
"\n",
"\n",
"# Append classifier to preprocessing pipeline.\n",
"# Now we have a full prediction pipeline.\n",
"clf = Pipeline(steps=[('preprocessor', DataFrameMapper(transformations)),\n",
" ('classifier', LogisticRegression(solver='lbfgs'))])\n",
"'''"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a Logistic Regression model, which you want to explain"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Append classifier to preprocessing pipeline.\n",
"# Now we have a full prediction pipeline.\n",
"clf = Pipeline(steps=[('preprocessor', transformations),\n",
" ('classifier', LogisticRegression(solver='lbfgs'))])\n",
"model = clf.fit(x_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain predictions on your local machine"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# 1. Using SHAP TabularExplainer\n",
"# When the last parameter allow_all_transformations is passed, we handle many to one and many to many transformations to \n",
"# generate approximations to raw feature importances. When this flag is passed, for transformations not recognized as one to \n",
"# many, we distribute feature importances evenly to raw features generating them.\n",
"# clf.steps[-1][1] returns the trained classification model\n",
"explainer = TabularExplainer(clf.steps[-1][1], \n",
" initialization_examples=x_train, \n",
" features=x_train.columns, \n",
" transformations=transformations, \n",
" allow_all_transformations=True)\n",
"\n",
"\n",
"\n",
"\n",
"# 2. Using MimicExplainer\n",
"# augment_data is optional and if true, oversamples the initialization examples to improve surrogate model accuracy to fit original model. Useful for high-dimensional data where the number of rows is less than the number of columns. \n",
"# max_num_of_augmentations is optional and defines max number of times we can increase the input data size.\n",
"# LGBMExplainableModel can be replaced with LinearExplainableModel, SGDExplainableModel, or DecisionTreeExplainableModel\n",
"# explainer = MimicExplainer(clf.steps[-1][1], \n",
"# x_train, \n",
"# LGBMExplainableModel, \n",
"# augment_data=True, \n",
"# max_num_of_augmentations=10, \n",
"# features=x_train.columns, \n",
"# transformations=transformations, \n",
"# allow_all_transformations=True)\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"# 3. Using PFIExplainer\n",
"\n",
"# Use the parameter \"metric\" to pass a metric name or function to evaluate the permutation. \n",
"# Note that if a metric function is provided a higher value must be better.\n",
"# Otherwise, take the negative of the function or set the parameter \"is_error_metric\" to True.\n",
"# Default metrics: \n",
"# F1 Score for binary classification, F1 Score with micro average for multiclass classification and\n",
"# Mean absolute error for regression\n",
"\n",
"\n",
"# explainer = PFIExplainer(clf.steps[-1][1], \n",
"# features=x_train.columns, \n",
"# transformations=transformations)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate global explanations\n",
"Explain overall model predictions (global explanation)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Passing in test dataset for evaluation examples - note it must be a representative sample of the original data\n",
"# x_train can be passed as well, but with more examples explanations will take longer although they may be more accurate\n",
"\n",
"global_explanation = explainer.explain_global(x_test)\n",
"\n",
"# Note: if you used the PFIExplainer in the previous step, use the next line of code instead\n",
"# global_explanation = explainer.explain_global(x_test, true_labels=y_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Sorted SHAP values\n",
"print('ranked global importance values: {}'.format(global_explanation.get_ranked_global_values()))\n",
"# Corresponding feature names\n",
"print('ranked global importance names: {}'.format(global_explanation.get_ranked_global_names()))\n",
"# Feature ranks (based on original order of features)\n",
"print('global importance rank: {}'.format(global_explanation.global_importance_rank))\n",
"# Per class feature names\n",
"print('ranked per class feature names: {}'.format(global_explanation.get_ranked_per_class_names()))\n",
"# Per class feature importance values\n",
"print('ranked per class feature values: {}'.format(global_explanation.get_ranked_per_class_values()))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Print out a dictionary that holds the sorted feature importance names and values\n",
"print('global importance rank: {}'.format(global_explanation.get_feature_importance_dict()))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain overall model predictions as a collection of local (instance-level) explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# feature shap values for all features and all data points in the training data\n",
"print('local importance values: {}'.format(global_explanation.local_importance_values))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate local explanations\n",
"Explain local data points (individual instances)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Note: PFIExplainer does not support local explanations\n",
"# You can pass a specific data point or a group of data points to the explain_local function\n",
"\n",
"# E.g., Explain the first data point in the test set\n",
"instance_num = 1\n",
"local_explanation = explainer.explain_local(x_test[:instance_num])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Get the prediction for the first member of the test set and explain why model made that prediction\n",
"prediction_value = clf.predict(x_test)[instance_num]\n",
"\n",
"sorted_local_importance_values = local_explanation.get_ranked_local_values()[prediction_value]\n",
"sorted_local_importance_names = local_explanation.get_ranked_local_names()[prediction_value]\n",
"\n",
"print('local importance values: {}'.format(sorted_local_importance_values))\n",
"print('local importance names: {}'.format(sorted_local_importance_names))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Load the visualization dashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(global_explanation, model, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
" \n",
"1. [Training time: regression problem](./explain-regression-local.ipynb)\n",
"1. [Training time: binary classification problem](./explain-binary-classification-local.ipynb)\n",
"1. [Training time: multiclass classification problem](./explain-multiclass-classification-local.ipynb)\n",
"1. [Explain models with simple feature transformations](./simple-feature-transformations-explain-local.ipynb)\n",
"1. [Save model explanations via Azure Machine Learning Run History](../azure-integration/run-history/save-retrieve-explanations-run-history.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../azure-integration/remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. Inferencing time: deploy a classification model and explainer:\n",
" 1. [Deploy a locally-trained model and explainer](../azure-integration/scoring-time/train-explain-model-locally-and-deploy.ipynb)\n",
" 1. [Deploy a remotely-trained model and explainer](../azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.ipynb)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

7
how-to-use-azureml/explain-model/tabular-data/advanced-feature-transformations-explain-local.yml Normal file
View File

@@ -0,0 +1,7 @@
name: advanced-feature-transformations-explain-local
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model
- sklearn-pandas

404
how-to-use-azureml/explain-model/tabular-data/explain-binary-classification-local.ipynb Normal file
View File

@@ -0,0 +1,404 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/tabular-data/explain-binary-classification-local.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Explain binary classification model predictions\n",
"_**This notebook showcases how to use the Azure Machine Learning Interpretability SDK to explain and visualize a binary classification model predictions.**_\n",
"\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Train a binary classification model\n",
" 1. Explain the model\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Visualize results](#Visualize)\n",
"1. [Next steps](#Next)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"This notebook illustrates how to explain a binary classification model predictions locally at training time without contacting any Azure services.\n",
"It demonstrates the API calls that you need to make to get the global and local explanations and a visualization dashboard that provides an interactive way of discovering patterns in data and explanations.\n",
"\n",
"We will showcase three tabular data explainers: TabularExplainer (SHAP), MimicExplainer (global surrogate), and PFIExplainer.\n",
"\n",
"| ![Interpretability Toolkit Architecture](./img/interpretability-architecture.PNG) |\n",
"|:--:|\n",
"| *Interpretability Toolkit Architecture* |\n",
"\n",
"Problem: Breast cancer diagnosis classification with scikit-learn (run model explainer locally)\n",
"\n",
"1. Train a SVM classification model using Scikit-learn\n",
"2. Run 'explain_model' globally and locally with full dataset in local mode, which doesn't contact any Azure services.\n",
"3. Visualize the global and local explanations with the visualization dashboard.\n",
"---\n",
"\n",
"## Setup\n",
"\n",
"You will need to have extensions enabled prior to jupyter kernel starting to see the visualization dashboard.\n",
"```\n",
"(myenv) $ jupyter nbextension install --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"(myenv) $ jupyter nbextension enable --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"```\n",
"Or\n",
"\n",
"```\n",
"(myenv) $ jupyter nbextension install azureml.contrib.explain.model.visualize --user --py\n",
"(myenv) $ jupyter nbextension enable azureml.contrib.explain.model.visualize --user --py\n",
"```\n",
"\n",
"If you are using Jupyter Labs run the following commands instead:\n",
"```\n",
"(myenv) $ jupyter labextension install @jupyter-widgets/jupyterlab-manager\n",
"(myenv) $ jupyter labextension install microsoft-mli-widget\n",
"```\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"\n",
"### Run model explainer locally at training time"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import load_breast_cancer\n",
"from sklearn import svm\n",
"\n",
"# Explainers:\n",
"# 1. SHAP Tabular Explainer\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"\n",
"# OR\n",
"\n",
"# 2. Mimic Explainer\n",
"from azureml.explain.model.mimic.mimic_explainer import MimicExplainer\n",
"# You can use one of the following four interpretable models as a global surrogate to the black box model\n",
"from azureml.explain.model.mimic.models.lightgbm_model import LGBMExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import LinearExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import SGDExplainableModel\n",
"from azureml.explain.model.mimic.models.tree_model import DecisionTreeExplainableModel\n",
"\n",
"# OR\n",
"\n",
"# 3. PFI Explainer\n",
"from azureml.explain.model.permutation.permutation_importance import PFIExplainer "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load the breast cancer diagnosis data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"breast_cancer_data = load_breast_cancer()\n",
"classes = breast_cancer_data.target_names.tolist()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Split data into train and test\n",
"from sklearn.model_selection import train_test_split\n",
"x_train, x_test, y_train, y_test = train_test_split(breast_cancer_data.data, breast_cancer_data.target, test_size=0.2, random_state=0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a SVM classification model, which you want to explain"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"clf = svm.SVC(gamma=0.001, C=100., probability=True)\n",
"model = clf.fit(x_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain predictions on your local machine"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# 1. Using SHAP TabularExplainer\n",
"explainer = TabularExplainer(model, \n",
" x_train, \n",
" features=breast_cancer_data.feature_names, \n",
" classes=classes)\n",
"\n",
"\n",
"\n",
"\n",
"# 2. Using MimicExplainer\n",
"# augment_data is optional and if true, oversamples the initialization examples to improve surrogate model accuracy to fit original model. Useful for high-dimensional data where the number of rows is less than the number of columns. \n",
"# max_num_of_augmentations is optional and defines max number of times we can increase the input data size.\n",
"# LGBMExplainableModel can be replaced with LinearExplainableModel, SGDExplainableModel, or DecisionTreeExplainableModel\n",
"# explainer = MimicExplainer(model, \n",
"# x_train, \n",
"# LGBMExplainableModel, \n",
"# augment_data=True, \n",
"# max_num_of_augmentations=10, \n",
"# features=breast_cancer_data.feature_names, \n",
"# classes=classes)\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"# 3. Using PFIExplainer\n",
"\n",
"# Use the parameter \"metric\" to pass a metric name or function to evaluate the permutation. \n",
"# Note that if a metric function is provided a higher value must be better.\n",
"# Otherwise, take the negative of the function or set the parameter \"is_error_metric\" to True.\n",
"# Default metrics: \n",
"# F1 Score for binary classification, F1 Score with micro average for multiclass classification and\n",
"# Mean absolute error for regression\n",
"\n",
"# explainer = PFIExplainer(model, \n",
"# features=breast_cancer_data.feature_names, \n",
"# classes=classes)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate global explanations\n",
"Explain overall model predictions (global explanation)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Passing in test dataset for evaluation examples - note it must be a representative sample of the original data\n",
"# x_train can be passed as well, but with more examples explanations will take longer although they may be more accurate\n",
"global_explanation = explainer.explain_global(x_test)\n",
"\n",
"# Note: if you used the PFIExplainer in the previous step, use the next line of code instead\n",
"# global_explanation = explainer.explain_global(x_test, true_labels=y_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Sorted SHAP values\n",
"print('ranked global importance values: {}'.format(global_explanation.get_ranked_global_values()))\n",
"# Corresponding feature names\n",
"print('ranked global importance names: {}'.format(global_explanation.get_ranked_global_names()))\n",
"# Feature ranks (based on original order of features)\n",
"print('global importance rank: {}'.format(global_explanation.global_importance_rank))\n",
"\n",
"# Note: PFIExplainer does not support per class explanations\n",
"# Per class feature names\n",
"print('ranked per class feature names: {}'.format(global_explanation.get_ranked_per_class_names()))\n",
"# Per class feature importance values\n",
"print('ranked per class feature values: {}'.format(global_explanation.get_ranked_per_class_values()))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Print out a dictionary that holds the sorted feature importance names and values\n",
"print('global importance rank: {}'.format(global_explanation.get_feature_importance_dict()))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain overall model predictions as a collection of local (instance-level) explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# feature shap values for all features and all data points in the training data\n",
"print('local importance values: {}'.format(global_explanation.local_importance_values))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate local explanations\n",
"Explain local data points (individual instances)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Note: PFIExplainer does not support local explanations\n",
"# You can pass a specific data point or a group of data points to the explain_local function\n",
"\n",
"# E.g., Explain the first data point in the test set\n",
"instance_num = 0\n",
"local_explanation = explainer.explain_local(x_test[instance_num,:])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Get the prediction for the first member of the test set and explain why model made that prediction\n",
"prediction_value = clf.predict(x_test)[instance_num]\n",
"\n",
"sorted_local_importance_values = local_explanation.get_ranked_local_values()[prediction_value]\n",
"sorted_local_importance_names = local_explanation.get_ranked_local_names()[prediction_value]\n",
"\n",
"print('local importance values: {}'.format(sorted_local_importance_values))\n",
"print('local importance names: {}'.format(sorted_local_importance_names))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Load the visualization dashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(global_explanation, model, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
" \n",
"1. [Training time: regression problem](./explain-regression-local.ipynb)\n",
"1. [Training time: multiclass classification problem](./explain-multiclass-classification-local.ipynb)\n",
"1. Explain models with engineered features:\n",
" 1. [Simple feature transformations](./simple-feature-transformations-explain-local.ipynb)\n",
" 1. [Advanced feature transformations](./advanced-feature-transformations-explain-local.ipynb)\n",
"1. [Save model explanations via Azure Machine Learning Run History](../azure-integration/run-history/save-retrieve-explanations-run-history.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../azure-integration/remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. Inferencing time: deploy a classification model and explainer:\n",
" 1. [Deploy a locally-trained model and explainer](../azure-integration/scoring-time/train-explain-model-locally-and-deploy.ipynb)\n",
" 1. [Deploy a remotely-trained model and explainer](../azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.ipynb)\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

6
how-to-use-azureml/explain-model/tabular-data/explain-binary-classification-local.yml Normal file
View File

@@ -0,0 +1,6 @@
name: explain-binary-classification-local
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model

402
how-to-use-azureml/explain-model/tabular-data/explain-multiclass-classification-local.ipynb Normal file
View File

@@ -0,0 +1,402 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/tabular-data/explain-multiclass-classification-local.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Explain multiclass classification model's predictions\n",
"_**This notebook showcases how to use the Azure Machine Learning Interpretability SDK to explain and visualize a multiclass classification model predictions.**_\n",
"\n",
"\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Train a multiclass classification model\n",
" 1. Explain the model\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Visualize results](#Visualize)\n",
"1. [Next steps](#Next)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"This notebook illustrates how to explain a multiclass classification model predictions locally at training time without contacting any Azure services.\n",
"It demonstrates the API calls that you need to make to get the global and local explanations and a visualization dashboard that provides an interactive way of discovering patterns in data and explanations.\n",
"\n",
"We will showcase three tabular data explainers: TabularExplainer (SHAP), MimicExplainer (global surrogate), and PFIExplainer.\n",
"\n",
"| ![Interpretability Toolkit Architecture](./img/interpretability-architecture.PNG) |\n",
"|:--:|\n",
"| *Interpretability Toolkit Architecture* |\n",
"\n",
"Problem: Iris flower classification with scikit-learn (run model explainer locally)\n",
"\n",
"1. Train a SVM classification model using Scikit-learn\n",
"2. Run 'explain_model' globally and locally with full dataset in local mode, which doesn't contact any Azure services.\n",
"3. Visualize the global and local explanations with the visualization dashboard.\n",
"---\n",
"\n",
"## Setup\n",
"\n",
"You will need to have extensions enabled prior to jupyter kernel starting to see the visualization dashboard.\n",
"```\n",
"(myenv) $ jupyter nbextension install --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"(myenv) $ jupyter nbextension enable --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"```\n",
"Or\n",
"\n",
"```\n",
"(myenv) $ jupyter nbextension install azureml.contrib.explain.model.visualize --user --py\n",
"(myenv) $ jupyter nbextension enable azureml.contrib.explain.model.visualize --user --py\n",
"```\n",
"\n",
"If you are using Jupyter Labs run the following commands instead:\n",
"```\n",
"(myenv) $ jupyter labextension install @jupyter-widgets/jupyterlab-manager\n",
"(myenv) $ jupyter labextension install microsoft-mli-widget\n",
"```\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"\n",
"### Run model explainer locally at training time"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.datasets import load_iris\n",
"from sklearn import svm\n",
"\n",
"# Explainers:\n",
"# 1. SHAP Tabular Explainer\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"\n",
"# OR\n",
"\n",
"# 2. Mimic Explainer\n",
"from azureml.explain.model.mimic.mimic_explainer import MimicExplainer\n",
"# You can use one of the following four interpretable models as a global surrogate to the black box model\n",
"from azureml.explain.model.mimic.models.lightgbm_model import LGBMExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import LinearExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import SGDExplainableModel\n",
"from azureml.explain.model.mimic.models.tree_model import DecisionTreeExplainableModel\n",
"\n",
"# OR\n",
"\n",
"# 3. PFI Explainer\n",
"from azureml.explain.model.permutation.permutation_importance import PFIExplainer "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load the Iris flower dataset"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"iris = load_iris()\n",
"X = iris['data']\n",
"y = iris['target']\n",
"classes = iris['target_names']\n",
"feature_names = iris['feature_names']"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Split data into train and test\n",
"from sklearn.model_selection import train_test_split\n",
"x_train, x_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a SVM classification model, which you want to explain"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"clf = svm.SVC(gamma=0.001, C=100., probability=True)\n",
"model = clf.fit(x_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain predictions on your local machine"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# 1. Using SHAP TabularExplainer\n",
"explainer = TabularExplainer(model, \n",
" x_train, \n",
" features=feature_names, \n",
" classes=classes)\n",
"\n",
"\n",
"\n",
"\n",
"# 2. Using MimicExplainer\n",
"# augment_data is optional and if true, oversamples the initialization examples to improve surrogate model accuracy to fit original model. Useful for high-dimensional data where the number of rows is less than the number of columns. \n",
"# max_num_of_augmentations is optional and defines max number of times we can increase the input data size.\n",
"# LGBMExplainableModel can be replaced with LinearExplainableModel, SGDExplainableModel, or DecisionTreeExplainableModel\n",
"# explainer = MimicExplainer(model, \n",
"# x_train, \n",
"# LGBMExplainableModel, \n",
"# augment_data=True, \n",
"# max_num_of_augmentations=10, \n",
"# features=feature_names, \n",
"# classes=classes)\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"# 3. Using PFIExplainer\n",
"\n",
"# Use the parameter \"metric\" to pass a metric name or function to evaluate the permutation. \n",
"# Note that if a metric function is provided a higher value must be better.\n",
"# Otherwise, take the negative of the function or set the parameter \"is_error_metric\" to True.\n",
"# Default metrics: \n",
"# F1 Score for binary classification, F1 Score with micro average for multiclass classification and\n",
"# Mean absolute error for regression\n",
"\n",
"# explainer = PFIExplainer(model, \n",
"# features=feature_names, \n",
"# classes=classes)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate global explanations\n",
"Explain overall model predictions (global explanation)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Passing in test dataset for evaluation examples - note it must be a representative sample of the original data\n",
"# x_train can be passed as well, but with more examples explanations will take longer although they may be more accurate\n",
"global_explanation = explainer.explain_global(x_test)\n",
"\n",
"# Note: if you used the PFIExplainer in the previous step, use the next line of code instead\n",
"# global_explanation = explainer.explain_global(x_test, true_labels=y_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Sorted SHAP values\n",
"print('ranked global importance values: {}'.format(global_explanation.get_ranked_global_values()))\n",
"# Corresponding feature names\n",
"print('ranked global importance names: {}'.format(global_explanation.get_ranked_global_names()))\n",
"# Feature ranks (based on original order of features)\n",
"print('global importance rank: {}'.format(global_explanation.global_importance_rank))\n",
"\n",
"# Note: PFIExplainer does not support per class explanations\n",
"# Per class feature names\n",
"print('ranked per class feature names: {}'.format(global_explanation.get_ranked_per_class_names()))\n",
"# Per class feature importance values\n",
"print('ranked per class feature values: {}'.format(global_explanation.get_ranked_per_class_values()))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Print out a dictionary that holds the sorted feature importance names and values\n",
"print('global importance rank: {}'.format(global_explanation.get_feature_importance_dict()))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain overall model predictions as a collection of local (instance-level) explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# feature shap values for all features and all data points in the training data\n",
"print('local importance values: {}'.format(global_explanation.local_importance_values))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate local explanations\n",
"Explain local data points (individual instances)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Note: PFIExplainer does not support local explanations\n",
"# You can pass a specific data point or a group of data points to the explain_local function\n",
"\n",
"# E.g., Explain the first data point in the test set\n",
"instance_num = 0\n",
"local_explanation = explainer.explain_local(x_test[instance_num,:])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Get the prediction for the first member of the test set and explain why model made that prediction\n",
"prediction_value = clf.predict(x_test)[instance_num]\n",
"\n",
"sorted_local_importance_values = local_explanation.get_ranked_local_values()[prediction_value]\n",
"sorted_local_importance_names = local_explanation.get_ranked_local_names()[prediction_value]\n",
"\n",
"print('local importance values: {}'.format(sorted_local_importance_values))\n",
"print('local importance names: {}'.format(sorted_local_importance_names))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Load the visualization dashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(global_explanation, model, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
"\n",
"1. [Training time: regression problem](./explain-regression-local.ipynb) \n",
"1. [Training time: binary classification problem](./explain-binary-classification-local.ipynb)\n",
"1. Explain models with engineered features:\n",
" 1. [Simple feature transformations](./simple-feature-transformations-explain-local.ipynb)\n",
" 1. [Advanced feature transformations](./advanced-feature-transformations-explain-local.ipynb)\n",
"1. [Save model explanations via Azure Machine Learning Run History](../azure-integration/run-history/save-retrieve-explanations-run-history.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../azure-integration/remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. Inferencing time: deploy a classification model and explainer:\n",
" 1. [Deploy a locally-trained model and explainer](../azure-integration/scoring-time/train-explain-model-locally-and-deploy.ipynb)\n",
" 1. [Deploy a remotely-trained model and explainer](../azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.ipynb)\n",
"\u00e2\u20ac\u2039\n"
]
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

6
how-to-use-azureml/explain-model/tabular-data/explain-multiclass-classification-local.yml Normal file
View File

@@ -0,0 +1,6 @@
name: explain-multiclass-classification-local
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model

397
how-to-use-azureml/explain-model/tabular-data/explain-regression-local.ipynb Normal file
View File

@@ -0,0 +1,397 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/tabular-data/explain-regression-local.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Explain regression model predictions\n",
"_**This notebook showcases how to use the Azure Machine Learning Interpretability SDK to explain and visualize a regression model predictions.**_\n",
"\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Train a regressor model\n",
" 1. Explain the model\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Visualize results](#Visualize)\n",
"1. [Next steps](#Next)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"This notebook illustrates how to explain regression model predictions locally at training time without contacting any Azure services.\n",
"It demonstrates the API calls that you need to make to get the global and local explanations and a visualization dashboard that provides an interactive way of discovering patterns in data and explanations.\n",
"\n",
"We will showcase three tabular data explainers: TabularExplainer (SHAP), MimicExplainer (global surrogate), and PFIExplainer.\n",
"\n",
"| ![Interpretability Toolkit Architecture](./img/interpretability-architecture.PNG) |\n",
"|:--:|\n",
"| *Interpretability Toolkit Architecture* |\n",
"\n",
"Problem: Boston Housing Price Prediction with scikit-learn (run model explainer locally)\n",
"\n",
"1. Train a GradientBoosting regression model using Scikit-learn\n",
"2. Run 'explain_model' globally and locally with full dataset in local mode, which doesn't contact any Azure services.\n",
"3. Visualize the global and local explanations with the visualization dashboard.\n",
"---\n",
"\n",
"## Setup\n",
"\n",
"You will need to have extensions enabled prior to jupyter kernel starting to see the visualization dashboard.\n",
"```\n",
"(myenv) $ jupyter nbextension install --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"(myenv) $ jupyter nbextension enable --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"```\n",
"Or\n",
"\n",
"```\n",
"(myenv) $ jupyter nbextension install azureml.contrib.explain.model.visualize --user --py\n",
"(myenv) $ jupyter nbextension enable azureml.contrib.explain.model.visualize --user --py\n",
"```\n",
"\n",
"If you are using Jupyter Labs run the following commands instead:\n",
"```\n",
"(myenv) $ jupyter labextension install @jupyter-widgets/jupyterlab-manager\n",
"(myenv) $ jupyter labextension install microsoft-mli-widget\n",
"```\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"\n",
"### Run model explainer locally at training time"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn import datasets\n",
"from sklearn.ensemble import GradientBoostingRegressor\n",
"\n",
"# Explainers:\n",
"# 1. SHAP Tabular Explainer\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"\n",
"# OR\n",
"\n",
"# 2. Mimic Explainer\n",
"from azureml.explain.model.mimic.mimic_explainer import MimicExplainer\n",
"# You can use one of the following four interpretable models as a global surrogate to the black box model\n",
"from azureml.explain.model.mimic.models.lightgbm_model import LGBMExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import LinearExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import SGDExplainableModel\n",
"from azureml.explain.model.mimic.models.tree_model import DecisionTreeExplainableModel\n",
"\n",
"# OR\n",
"\n",
"# 3. PFI Explainer\n",
"from azureml.explain.model.permutation.permutation_importance import PFIExplainer "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load the Boston house price data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"boston_data = datasets.load_boston()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Split data into train and test\n",
"from sklearn.model_selection import train_test_split\n",
"x_train, x_test, y_train, y_test = train_test_split(boston_data.data, boston_data.target, test_size=0.2, random_state=0)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a GradientBoosting regression model, which you want to explain"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"reg = GradientBoostingRegressor(n_estimators=100, max_depth=4,\n",
" learning_rate=0.1, loss='huber',\n",
" random_state=1)\n",
"model = reg.fit(x_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain predictions on your local machine"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# 1. Using SHAP TabularExplainer\n",
"explainer = TabularExplainer(model, \n",
" x_train, \n",
" features = boston_data.feature_names)\n",
"\n",
"\n",
"\n",
"\n",
"# 2. Using MimicExplainer\n",
"# augment_data is optional and if true, oversamples the initialization examples to improve surrogate model accuracy to fit original model. Useful for high-dimensional data where the number of rows is less than the number of columns. \n",
"# max_num_of_augmentations is optional and defines max number of times we can increase the input data size.\n",
"# LGBMExplainableModel can be replaced with LinearExplainableModel, SGDExplainableModel, or DecisionTreeExplainableModel\n",
"# explainer = MimicExplainer(model, \n",
"# x_train, \n",
"# LGBMExplainableModel, \n",
"# augment_data=True, \n",
"# max_num_of_augmentations=10, \n",
"# features=boston_data.feature_names)\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"# 3. Using PFIExplainer\n",
"\n",
"# Use the parameter \"metric\" to pass a metric name or function to evaluate the permutation. \n",
"# Note that if a metric function is provided a higher value must be better.\n",
"# Otherwise, take the negative of the function or set the parameter \"is_error_metric\" to True.\n",
"# Default metrics: \n",
"# F1 Score for binary classification, F1 Score with micro average for multiclass classification and\n",
"# Mean absolute error for regression\n",
"\n",
"# explainer = PFIExplainer(model, \n",
"# features=boston_data.feature_names)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate global explanations\n",
"Explain overall model predictions (global explanation)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Passing in test dataset for evaluation examples - note it must be a representative sample of the original data\n",
"# x_train can be passed as well, but with more examples explanations will take longer although they may be more accurate\n",
"global_explanation = explainer.explain_global(x_test)\n",
"\n",
"# Note: if you used the PFIExplainer in the previous step, use the next line of code instead\n",
"# global_explanation = explainer.explain_global(x_test, true_labels=y_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Sorted SHAP values \n",
"print('ranked global importance values: {}'.format(global_explanation.get_ranked_global_values()))\n",
"# Corresponding feature names\n",
"print('ranked global importance names: {}'.format(global_explanation.get_ranked_global_names()))\n",
"# Feature ranks (based on original order of features)\n",
"print('global importance rank: {}'.format(global_explanation.global_importance_rank))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Print out a dictionary that holds the sorted feature importance names and values\n",
"print('global importance rank: {}'.format(global_explanation.get_feature_importance_dict()))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain overall model predictions as a collection of local (instance-level) explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Note: PFIExplainer does not support local explanations\n",
"# feature shap values for all features and all data points in the training data\n",
"print('local importance values: {}'.format(global_explanation.local_importance_values))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate local explanations\n",
"Explain local data points (individual instances)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Note: PFIExplainer does not support local explanations\n",
"# You can pass a specific data point or a group of data points to the explain_local function\n",
"\n",
"# E.g., Explain the first data point in the test set\n",
"local_explanation = explainer.explain_local(x_test[0,:])\n",
"\n",
"# E.g., Explain the first five data points in the test set\n",
"# local_explanation_group = explainer.explain_local(x_test[0:4,:])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Sorted local feature importance information; reflects the original feature order\n",
"sorted_local_importance_names = local_explanation.get_ranked_local_names()\n",
"sorted_local_importance_values = local_explanation.get_ranked_local_values()\n",
"\n",
"print('sorted local importance names: {}'.format(sorted_local_importance_names))\n",
"print('sorted local importance values: {}'.format(sorted_local_importance_values))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Load the visualization dashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(global_explanation, model, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
" \n",
"1. [Training time: binary classification problem](./explain-binary-classification-local.ipynb)\n",
"1. [Training time: multiclass classification problem](./explain-multiclass-classification-local.ipynb)\n",
"1. Explain models with engineered features:\n",
" 1. [Simple feature transformations](./simple-feature-transformations-explain-local.ipynb)\n",
" 1. [Advanced feature transformations](./advanced-feature-transformations-explain-local.ipynb)\n",
"1. [Save model explanations via Azure Machine Learning Run History](../azure-integration/run-history/save-retrieve-explanations-run-history.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../azure-integration/remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. Inferencing time: deploy a classification model and explainer:\n",
" 1. [Deploy a locally-trained model and explainer](../azure-integration/scoring-time/train-explain-model-locally-and-deploy.ipynb)\n",
" 1. [Deploy a remotely-trained model and explainer](../azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.ipynb)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

6
how-to-use-azureml/explain-model/tabular-data/explain-regression-local.yml Normal file
View File

@@ -0,0 +1,6 @@
name: explain-regression-local
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model

531
how-to-use-azureml/explain-model/tabular-data/simple-feature-transformations-explain-local.ipynb Normal file
View File

@@ -0,0 +1,531 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/explain-model/tabular-data/simple-feature-transformations-explain-local.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Explain binary classification model predictions with raw feature transformations\n",
"_**This notebook showcases how to use the Azure Machine Learning Interpretability SDK to explain and visualize a binary classification model that uses one to one and one to many feature transformations.**_\n",
"\n",
"\n",
"## Table of Contents\n",
"\n",
"1. [Introduction](#Introduction)\n",
"1. [Setup](#Setup)\n",
"1. [Run model explainer locally at training time](#Explain)\n",
" 1. Apply feature transformations\n",
" 1. Train a binary classification model\n",
" 1. Explain the model on raw features\n",
" 1. Generate global explanations\n",
" 1. Generate local explanations\n",
"1. [Visualize results](#Visualize)\n",
"1. [Next steps](#Next%20steps)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Introduction\n",
"\n",
"This notebook illustrates creating explanations for a binary classification model, IBM employee attrition classification, that uses one to one and one to many feature transformations from raw data to engineered features. The one to many feature transformations include one hot encoding on categorical features. The one to one feature transformations apply standard scaling on numeric features. Our tabular data explainer is then used to get raw feature importances.\n",
"\n",
"\n",
"We will showcase raw feature transformations with three tabular data explainers: TabularExplainer (SHAP), MimicExplainer (global surrogate), and PFIExplainer.\n",
"\n",
"| ![Interpretability Toolkit Architecture](./img/interpretability-architecture.PNG) |\n",
"|:--:|\n",
"| *Interpretability Toolkit Architecture* |\n",
"\n",
"Problem: IBM employee attrition classification with scikit-learn (run model explainer locally)\n",
"\n",
"1. Transform raw features to engineered features\n",
"2. Train a SVC classification model using Scikit-learn\n",
"3. Run 'explain_model' globally and locally with full dataset in local mode, which doesn't contact any Azure services.\n",
"4. Visualize the global and local explanations with the visualization dashboard.\n",
"---\n",
"\n",
"## Setup\n",
"\n",
"You will need to have extensions enabled prior to jupyter kernel starting to see the visualization dashboard.\n",
"```\n",
"(myenv) $ jupyter nbextension install --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"(myenv) $ jupyter nbextension enable --py --sys-prefix azureml.contrib.explain.model.visualize\n",
"```\n",
"Or\n",
"\n",
"```\n",
"(myenv) $ jupyter nbextension install azureml.contrib.explain.model.visualize --user --py\n",
"(myenv) $ jupyter nbextension enable azureml.contrib.explain.model.visualize --user --py\n",
"```\n",
"\n",
"If you are using Jupyter Labs run the following commands instead:\n",
"```\n",
"(myenv) $ jupyter labextension install @jupyter-widgets/jupyterlab-manager\n",
"(myenv) $ jupyter labextension install microsoft-mli-widget\n",
"```\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Explain\n",
"\n",
"### Run model explainer locally at training time"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.pipeline import Pipeline\n",
"from sklearn.impute import SimpleImputer\n",
"from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
"from sklearn.svm import SVC\n",
"import pandas as pd\n",
"import numpy as np\n",
"\n",
"# Explainers:\n",
"# 1. SHAP Tabular Explainer\n",
"from azureml.explain.model.tabular_explainer import TabularExplainer\n",
"\n",
"# OR\n",
"\n",
"# 2. Mimic Explainer\n",
"from azureml.explain.model.mimic.mimic_explainer import MimicExplainer\n",
"# You can use one of the following four interpretable models as a global surrogate to the black box model\n",
"from azureml.explain.model.mimic.models.lightgbm_model import LGBMExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import LinearExplainableModel\n",
"from azureml.explain.model.mimic.models.linear_model import SGDExplainableModel\n",
"from azureml.explain.model.mimic.models.tree_model import DecisionTreeExplainableModel\n",
"\n",
"# OR\n",
"\n",
"# 3. PFI Explainer\n",
"from azureml.explain.model.permutation.permutation_importance import PFIExplainer "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Load the IBM employee attrition data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# get the IBM employee attrition dataset\n",
"outdirname = 'dataset.6.21.19'\n",
"try:\n",
" from urllib import urlretrieve\n",
"except ImportError:\n",
" from urllib.request import urlretrieve\n",
"import zipfile\n",
"zipfilename = outdirname + '.zip'\n",
"urlretrieve('https://publictestdatasets.blob.core.windows.net/data/' + zipfilename, zipfilename)\n",
"with zipfile.ZipFile(zipfilename, 'r') as unzip:\n",
" unzip.extractall('.')\n",
"attritionData = pd.read_csv('./WA_Fn-UseC_-HR-Employee-Attrition.csv')\n",
"\n",
"# Dropping Employee count as all values are 1 and hence attrition is independent of this feature\n",
"attritionData = attritionData.drop(['EmployeeCount'], axis=1)\n",
"# Dropping Employee Number since it is merely an identifier\n",
"attritionData = attritionData.drop(['EmployeeNumber'], axis=1)\n",
"\n",
"attritionData = attritionData.drop(['Over18'], axis=1)\n",
"\n",
"# Since all values are 80\n",
"attritionData = attritionData.drop(['StandardHours'], axis=1)\n",
"\n",
"# Converting target variables from string to numerical values\n",
"target_map = {'Yes': 1, 'No': 0}\n",
"attritionData[\"Attrition_numerical\"] = attritionData[\"Attrition\"].apply(lambda x: target_map[x])\n",
"target = attritionData[\"Attrition_numerical\"]\n",
"\n",
"attritionXData = attritionData.drop(['Attrition_numerical', 'Attrition'], axis=1)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Split data into train and test\n",
"from sklearn.model_selection import train_test_split\n",
"x_train, x_test, y_train, y_test = train_test_split(attritionXData, \n",
" target, \n",
" test_size = 0.2,\n",
" random_state=0,\n",
" stratify=target)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Creating dummy columns for each categorical feature\n",
"categorical = []\n",
"for col, value in attritionXData.iteritems():\n",
" if value.dtype == 'object':\n",
" categorical.append(col)\n",
" \n",
"# Store the numerical columns in a list numerical\n",
"numerical = attritionXData.columns.difference(categorical) "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Transform raw features"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can explain raw features by either using a `sklearn.compose.ColumnTransformer` or a list of fitted transformer tuples. The cell below uses `sklearn.compose.ColumnTransformer`. In case you want to run the example with the list of fitted transformer tuples, comment the cell below and uncomment the cell that follows after. "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.compose import ColumnTransformer\n",
"\n",
"# We create the preprocessing pipelines for both numeric and categorical data.\n",
"numeric_transformer = Pipeline(steps=[\n",
" ('imputer', SimpleImputer(strategy='median')),\n",
" ('scaler', StandardScaler())])\n",
"\n",
"categorical_transformer = Pipeline(steps=[\n",
" ('imputer', SimpleImputer(strategy='constant', fill_value='missing')),\n",
" ('onehot', OneHotEncoder(handle_unknown='ignore'))])\n",
"\n",
"transformations = ColumnTransformer(\n",
" transformers=[\n",
" ('num', numeric_transformer, numerical),\n",
" ('cat', categorical_transformer, categorical)])\n",
"\n",
"# Append classifier to preprocessing pipeline.\n",
"# Now we have a full prediction pipeline.\n",
"clf = Pipeline(steps=[('preprocessor', transformations),\n",
" ('classifier', SVC(kernel='linear', C = 1.0, probability=True))])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"'''\n",
"# Uncomment below if sklearn-pandas is not installed\n",
"#!pip install sklearn-pandas\n",
"from sklearn_pandas import DataFrameMapper\n",
"\n",
"# Impute, standardize the numeric features and one-hot encode the categorical features. \n",
"\n",
"\n",
"numeric_transformations = [([f], Pipeline(steps=[('imputer', SimpleImputer(strategy='median')), ('scaler', StandardScaler())])) for f in numerical]\n",
"\n",
"categorical_transformations = [([f], OneHotEncoder(handle_unknown='ignore', sparse=False)) for f in categorical]\n",
"\n",
"transformations = numeric_transformations + categorical_transformations\n",
"\n",
"# Append classifier to preprocessing pipeline.\n",
"# Now we have a full prediction pipeline.\n",
"clf = Pipeline(steps=[('preprocessor', transformations),\n",
" ('classifier', SVC(kernel='linear', C = 1.0, probability=True))]) \n",
"\n",
"\n",
"\n",
"'''"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Train a SVM classification model, which you want to explain"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model = clf.fit(x_train, y_train)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain predictions on your local machine"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# 1. Using SHAP TabularExplainer\n",
"# clf.steps[-1][1] returns the trained classification model\n",
"explainer = TabularExplainer(clf.steps[-1][1], \n",
" initialization_examples=x_train, \n",
" features=attritionXData.columns, \n",
" classes=[\"Not leaving\", \"leaving\"], \n",
" transformations=transformations)\n",
"\n",
"\n",
"\n",
"\n",
"# 2. Using MimicExplainer\n",
"# augment_data is optional and if true, oversamples the initialization examples to improve surrogate model accuracy to fit original model. Useful for high-dimensional data where the number of rows is less than the number of columns. \n",
"# max_num_of_augmentations is optional and defines max number of times we can increase the input data size.\n",
"# LGBMExplainableModel can be replaced with LinearExplainableModel, SGDExplainableModel, or DecisionTreeExplainableModel\n",
"# explainer = MimicExplainer(clf.steps[-1][1], \n",
"# x_train, \n",
"# LGBMExplainableModel, \n",
"# augment_data=True, \n",
"# max_num_of_augmentations=10, \n",
"# features=attritionXData.columns, \n",
"# classes=[\"Not leaving\", \"leaving\"], \n",
"# transformations=transformations)\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"# 3. Using PFIExplainer\n",
"\n",
"# Use the parameter \"metric\" to pass a metric name or function to evaluate the permutation. \n",
"# Note that if a metric function is provided a higher value must be better.\n",
"# Otherwise, take the negative of the function or set the parameter \"is_error_metric\" to True.\n",
"# Default metrics: \n",
"# F1 Score for binary classification, F1 Score with micro average for multiclass classification and\n",
"# Mean absolute error for regression\n",
"\n",
"# explainer = PFIExplainer(clf.steps[-1][1], \n",
"# features=x_train.columns, \n",
"# transformations=transformations,\n",
"# classes=[\"Not leaving\", \"leaving\"])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate global explanations\n",
"Explain overall model predictions (global explanation)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Passing in test dataset for evaluation examples - note it must be a representative sample of the original data\n",
"# x_train can be passed as well, but with more examples explanations will take longer although they may be more accurate\n",
"global_explanation = explainer.explain_global(x_test)\n",
"\n",
"# Note: if you used the PFIExplainer in the previous step, use the next line of code instead\n",
"# global_explanation = explainer.explain_global(x_test, true_labels=y_test)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Sorted SHAP values\n",
"print('ranked global importance values: {}'.format(global_explanation.get_ranked_global_values()))\n",
"# Corresponding feature names\n",
"print('ranked global importance names: {}'.format(global_explanation.get_ranked_global_names()))\n",
"# Feature ranks (based on original order of features)\n",
"print('global importance rank: {}'.format(global_explanation.global_importance_rank))\n",
"\n",
"# Note: PFIExplainer does not support per class explanations\n",
"# Per class feature names\n",
"print('ranked per class feature names: {}'.format(global_explanation.get_ranked_per_class_names()))\n",
"# Per class feature importance values\n",
"print('ranked per class feature values: {}'.format(global_explanation.get_ranked_per_class_values()))"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Print out a dictionary that holds the sorted feature importance names and values\n",
"print('global importance rank: {}'.format(global_explanation.get_feature_importance_dict()))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Explain overall model predictions as a collection of local (instance-level) explanations"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# feature shap values for all features and all data points in the training data\n",
"print('local importance values: {}'.format(global_explanation.local_importance_values))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Generate local explanations\n",
"Explain local data points (individual instances)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Note: PFIExplainer does not support local explanations\n",
"# You can pass a specific data point or a group of data points to the explain_local function\n",
"\n",
"# E.g., Explain the first data point in the test set\n",
"instance_num = 1\n",
"local_explanation = explainer.explain_local(x_test[:instance_num])"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Get the prediction for the first member of the test set and explain why model made that prediction\n",
"prediction_value = clf.predict(x_test)[instance_num]\n",
"\n",
"sorted_local_importance_values = local_explanation.get_ranked_local_values()[prediction_value]\n",
"sorted_local_importance_names = local_explanation.get_ranked_local_names()[prediction_value]\n",
"\n",
"print('local importance values: {}'.format(sorted_local_importance_values))\n",
"print('local importance names: {}'.format(sorted_local_importance_names))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Visualize\n",
"Load the visualization dashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.contrib.explain.model.visualize import ExplanationDashboard"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"ExplanationDashboard(global_explanation, model, x_test)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Next\n",
"Learn about other use cases of the explain package on a:\n",
" \n",
"1. [Training time: regression problem](./explain-regression-local.ipynb)\n",
"1. [Training time: binary classification problem](./explain-binary-classification-local.ipynb)\n",
"1. [Training time: multiclass classification problem](./explain-multiclass-classification-local.ipynb)\n",
"1. [Explain models with advanced feature transformations](./advanced-feature-transformations-explain-local.ipynb)\n",
"1. [Save model explanations via Azure Machine Learning Run History](../azure-integration/run-history/save-retrieve-explanations-run-history.ipynb)\n",
"1. [Run explainers remotely on Azure Machine Learning Compute (AMLCompute)](../azure-integration/remote-explanation/explain-model-on-amlcompute.ipynb)\n",
"1. Inferencing time: deploy a classification model and explainer:\n",
" 1. [Deploy a locally-trained model and explainer](../azure-integration/scoring-time/train-explain-model-locally-and-deploy.ipynb)\n",
" 1. [Deploy a remotely-trained model and explainer](../azure-integration/scoring-time/train-explain-model-on-amlcompute-and-deploy.ipynb)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"authors": [
{
"name": "mesameki"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

7
how-to-use-azureml/explain-model/tabular-data/simple-feature-transformations-explain-local.yml Normal file
View File

@@ -0,0 +1,7 @@
name: simple-feature-transformations-explain-local
dependencies:
- pip:
- azureml-sdk
- azureml-explain-model
- azureml-contrib-explain-model
- sklearn-pandas

2
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/README.md
View File

@@ -13,6 +13,8 @@ These notebooks below are designed to go in sequence.
8. [aml-pipelines-parameter-tuning-with-hyperdrive.ipynb](https://aka.ms/pl-hyperdrive): HyperDriveStep in Pipelines shows how you can do hyper parameter tuning using Pipelines.
9. [aml-pipelines-how-to-use-azurebatch-to-run-a-windows-executable.ipynb](https://aka.ms/pl-azbatch): AzureBatchStep can be used to run your custom code in AzureBatch cluster.
10. [aml-pipelines-setup-schedule-for-a-published-pipeline.ipynb](https://aka.ms/pl-schedule): Once you publish a Pipeline, you can schedule it to trigger based on an interval or on data change in a defined datastore.
11. [aml-pipelines-with-automated-machine-learning-step.ipynb](https://aka.ms/pl-automl): AutoMLStep in Pipelines shows how you can do automated machine learning using Pipelines.
12. [aml-pipelines-setup-versioned-pipeline-endpoints.ipynb](https://aka.ms/pl-ver-endpoint): This notebook shows how you can setup PipelineEndpoint and submit a Pipeline using the PipelineEndpoint.
![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/README.png)

236
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-data-transfer.ipynb
View File

@@ -22,9 +22,19 @@
"# Azure Machine Learning Pipeline with DataTranferStep\n",
"This notebook is used to demonstrate the use of DataTranferStep in Azure Machine Learning Pipeline.\n",
"\n",
"In certain cases, you will need to transfer data from one data location to another. For example, your data may be in Files storage and you may want to move it to Blob storage. Or, if your data is in an ADLS account and you want to make it available in the Blob storage. The built-in **DataTransferStep** class helps you transfer data in these situations.\n",
"In certain cases, you will need to transfer data from one data location to another. For example, your data may be in Azure SQL Database and you may want to move it to Azure Data Lake storage. Or, your data is in an ADLS account and you want to make it available in the Blob storage. The built-in **DataTransferStep** class helps you transfer data in these situations.\n",
"\n",
"The below example shows how to move data between an ADLS account, Blob storage, SQL Server, PostgreSQL server. "
"The below examples show how to move data between an ADLS account, Blob storage, SQL Server, PostgreSQL server. \n",
"\n",
"## Data transfer currently supports following storage types:\n",
"\n",
"| Data store | Supported as a source | Supported as a sink |\n",
"| --- | --- | --- |\n",
"| Azure Blob Storage | Yes | Yes |\n",
"| Azure Data Lake Storage Gen 1 | Yes | Yes |\n",
"| Azure Data Lake Storage Gen 2 | Yes | Yes |\n",
"| Azure SQL Database | Yes | Yes |\n",
"| Azure Database for PostgreSQL | Yes | No |"
]
},
{
@@ -62,8 +72,7 @@
"\n",
"Initialize a workspace object from persisted configuration. If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure the config file is present at .\\config.json\n",
"\n",
"If you don't have a config.json file, please go through the configuration Notebook located here:\n",
"https://github.com/Azure/MachineLearningNotebooks. \n",
"If you don't have a config.json file, please go through the [configuration Notebook](https://aka.ms/pl-config) first.\n",
"\n",
"This sets you up with a working config file that has information on your workspace, subscription id, etc. "
]
@@ -86,15 +95,58 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"## Register Datastores\n",
"\n",
"In the code cell below, you will need to fill in the appropriate values for the workspace name, datastore name, subscription id, resource group, store name, tenant id, client id, and client secret that are associated with your ADLS datastore. \n",
"## Register Datastores and create DataReferences\n",
"\n",
"For background on registering your data store, consult this article:\n",
"\n",
"https://docs.microsoft.com/en-us/azure/data-lake-store/data-lake-store-service-to-service-authenticate-using-active-directory\n",
"\n",
"### register datastores for Azure Data Lake and Azure Blob storage"
"> Please make sure to update the following code examples with appropriate values."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Azure Blob Storage"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from msrest.exceptions import HttpOperationError\n",
"\n",
"blob_datastore_name='MyBlobDatastore'\n",
"account_name=os.getenv(\"BLOB_ACCOUNTNAME_62\", \"<my-account-name>\") # Storage account name\n",
"container_name=os.getenv(\"BLOB_CONTAINER_62\", \"<my-container-name>\") # Name of Azure blob container\n",
"account_key=os.getenv(\"BLOB_ACCOUNT_KEY_62\", \"<my-account-key>\") # Storage account key\n",
"\n",
"try:\n",
" blob_datastore = Datastore.get(ws, blob_datastore_name)\n",
" print(\"found blob datastore with name: %s\" % blob_datastore_name)\n",
"except HttpOperationError:\n",
" blob_datastore = Datastore.register_azure_blob_container(\n",
" workspace=ws,\n",
" datastore_name=blob_datastore_name,\n",
" account_name=account_name, # Storage account name\n",
" container_name=container_name, # Name of Azure blob container\n",
" account_key=account_key) # Storage account key\n",
" print(\"registered blob datastore with name: %s\" % blob_datastore_name)\n",
"\n",
"blob_data_ref = DataReference(\n",
" datastore=blob_datastore,\n",
" data_reference_name=\"blob_test_data\",\n",
" path_on_datastore=\"testdata\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Azure Data Lake Storage Gen1"
]
},
{
@@ -128,34 +180,57 @@
" client_secret=client_secret) # the secret of service principal\n",
" print(\"registered datastore with name: %s\" % datastore_name)\n",
"\n",
"\n",
"\n",
"blob_datastore_name='MyBlobDatastore'\n",
"account_name=os.getenv(\"BLOB_ACCOUNTNAME_62\", \"<my-account-name>\") # Storage account name\n",
"container_name=os.getenv(\"BLOB_CONTAINER_62\", \"<my-container-name>\") # Name of Azure blob container\n",
"account_key=os.getenv(\"BLOB_ACCOUNT_KEY_62\", \"<my-account-key>\") # Storage account key\n",
"\n",
"try:\n",
" blob_datastore = Datastore.get(ws, blob_datastore_name)\n",
" print(\"found blob datastore with name: %s\" % blob_datastore_name)\n",
"except HttpOperationError:\n",
" blob_datastore = Datastore.register_azure_blob_container(\n",
" workspace=ws,\n",
" datastore_name=blob_datastore_name,\n",
" account_name=account_name, # Storage account name\n",
" container_name=container_name, # Name of Azure blob container\n",
" account_key=account_key) # Storage account key\"\n",
" print(\"registered blob datastore with name: %s\" % blob_datastore_name)\n",
"\n",
"# CLI:\n",
"# az ml datastore attach-blob -n <datastore-name> -a <account-name> -c <container-name> -k <account-key> [-t <sas-token>]"
"adls_data_ref = DataReference(\n",
" datastore=adls_datastore,\n",
" data_reference_name=\"adls_test_data\",\n",
" path_on_datastore=\"testdata\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### register datastores for Azure SQL Server and Azure database for PostgreSQL"
"### Azure Data Lake Storage Gen2"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"\n",
"adlsgen2_datastore_name = 'myadlsgen2datastore'\n",
"account_name=os.getenv(\"ADLSGEN2_ACCOUNTNAME_62\", \"<my-account-name>\") # ADLS Gen2 account name\n",
"tenant_id=os.getenv(\"ADLSGEN2_TENANT_62\", \"<my-tenant-id>\") # tenant id of service principal\n",
"client_id=os.getenv(\"ADLSGEN2_CLIENTID_62\", \"<my-client-id>\") # client id of service principal\n",
"client_secret=os.getenv(\"ADLSGEN2_CLIENT_SECRET_62\", \"<my-client-secret>\") # the secret of service principal\n",
"\n",
"try:\n",
" adlsgen2_datastore = Datastore.get(ws, adlsgen2_datastore_name)\n",
" print(\"found ADLS Gen2 datastore with name: %s\" % adlsgen2_datastore_name)\n",
"except:\n",
" adlsgen2_datastore = Datastore.register_azure_data_lake_gen2(\n",
" workspace=ws,\n",
" datastore_name=adlsgen2_datastore_name,\n",
" filesystem='test', # Name of ADLS Gen2 filesystem\n",
" account_name=account_name, # ADLS Gen2 account name\n",
" tenant_id=tenant_id, # tenant id of service principal\n",
" client_id=client_id, # client id of service principal\n",
" client_secret=client_secret) # the secret of service principal\n",
" print(\"registered datastore with name: %s\" % adlsgen2_datastore_name)\n",
"\n",
"adlsgen2_data_ref = DataReference(\n",
" datastore=adlsgen2_datastore,\n",
" data_reference_name='adlsgen2_test_data',\n",
" path_on_datastore='testdata')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Azure SQL Database"
]
},
{
@@ -186,6 +261,27 @@
" tenant_id=tenant_id)\n",
" print(\"registered sql databse datastore with name: %s\" % sql_datastore_name)\n",
"\n",
"from azureml.data.sql_data_reference import SqlDataReference\n",
"\n",
"sql_query_data_ref = SqlDataReference(\n",
" datastore=sql_datastore,\n",
" data_reference_name=\"sql_query_data_ref\",\n",
" sql_query=\"select top 1 * from TestData\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Azure Database for PostgreSQL"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"\n",
"psql_datastore_name=\"MyPostgreSqlDatastore\"\n",
"server_name=os.getenv(\"PSQL_SERVERNAME_62\", \"<my-server-name>\") # Name of PostgreSQL server \n",
@@ -205,73 +301,13 @@
" user_id=user_id,\n",
" user_password=user_password)\n",
" print(\"registered PostgreSQL databse datastore with name: %s\" % psql_datastore_name)\n",
" "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Create DataReferences\n",
"### create DataReferences for Azure Data Lake and Azure Blob storage"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"adls_datastore = Datastore(workspace=ws, name=\"MyAdlsDatastore\")\n",
"\n",
"# adls\n",
"adls_data_ref = DataReference(\n",
" datastore=adls_datastore,\n",
" data_reference_name=\"adls_test_data\",\n",
" path_on_datastore=\"testdata\")\n",
"\n",
"blob_datastore = Datastore(workspace=ws, name=\"MyBlobDatastore\")\n",
"\n",
"# blob data\n",
"blob_data_ref = DataReference(\n",
" datastore=blob_datastore,\n",
" data_reference_name=\"blob_test_data\",\n",
" path_on_datastore=\"testdata\")\n",
"\n",
"print(\"obtained adls, blob data references\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### create DataReferences for Azure SQL Server and Azure database for PostgreSQL"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.data.sql_data_reference import SqlDataReference\n",
"\n",
"sql_datastore = Datastore(workspace=ws, name=\"MySqlDatastore\")\n",
"\n",
"sql_query_data_ref = SqlDataReference(\n",
" datastore=sql_datastore,\n",
" data_reference_name=\"sql_query_data_ref\",\n",
" sql_query=\"select top 1 * from TestData\")\n",
"\n",
"\n",
"psql_datastore = Datastore(workspace=ws, name=\"MyPostgreSqlDatastore\")\n",
"\n",
"psql_query_data_ref = SqlDataReference(\n",
" datastore=psql_datastore,\n",
" data_reference_name=\"psql_query_data_ref\",\n",
" sql_query=\"SELECT * FROM testtable\")\n",
"\n",
"print(\"obtained Sql server, PostgreSQL data references\")"
" sql_query=\"SELECT * FROM testtable\")"
]
},
{
@@ -304,11 +340,7 @@
" \n",
"data_factory_compute = get_or_create_data_factory(ws, data_factory_name)\n",
"\n",
"print(\"setup data factory account complete\")\n",
"\n",
"# CLI:\n",
"# Create: az ml computetarget setup datafactory -n <name>\n",
"# BYOC: az ml computetarget attach datafactory -n <name> -i <resource-id>"
"print(\"setup data factory account complete\")"
]
},
{
@@ -357,6 +389,13 @@
"metadata": {},
"outputs": [],
"source": [
"\n",
"transfer_adlsgen2_to_blob = DataTransferStep(\n",
" name='transfer_adlsgen2_to_blob',\n",
" source_data_reference=adlsgen2_data_ref,\n",
" destination_data_reference=blob_data_ref,\n",
" compute_target=data_factory_compute)\n",
"\n",
"transfer_sql_to_blob = DataTransferStep(\n",
" name=\"transfer_sql_to_blob\",\n",
" source_data_reference=sql_query_data_ref,\n",
@@ -405,7 +444,7 @@
"pipeline_02 = Pipeline(\n",
" description=\"data_transfer_02\",\n",
" workspace=ws,\n",
" steps=[transfer_sql_to_blob,transfer_psql_to_blob])\n",
" steps=[transfer_sql_to_blob,transfer_psql_to_blob, transfer_adlsgen2_to_blob])\n",
"\n",
"pipeline_run_02 = Experiment(ws, \"Data_Transfer_example_02\").submit(pipeline_02)\n",
"pipeline_run_02.wait_for_completion()"
@@ -439,11 +478,12 @@
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"# Next: Databricks as a Compute Target\n",
"To use Databricks as a compute target from Azure Machine Learning Pipeline, a DatabricksStep is used. This [notebook](./aml-pipelines-use-databricks-as-compute-target.ipynb) demonstrates the use of a DatabricksStep in an Azure Machine Learning Pipeline."
"To use Databricks as a compute target from Azure Machine Learning Pipeline, a DatabricksStep is used. This [notebook](https://aka.ms/pl-databricks) demonstrates the use of a DatabricksStep in an Azure Machine Learning Pipeline."
]
}
],

8
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-getting-started.ipynb
View File

@@ -33,7 +33,7 @@
"\n",
"Azure's Machine Learning pipelines give you a way to combine multiple steps like these into one configurable workflow, so that multiple agents/users can share and/or reuse this workflow. Machine learning pipelines thus provide a consistent, reproducible mechanism for building, evaluating, deploying, and running ML systems.\n",
"\n",
"To get more information about Azure machine learning pipelines, please read our [Azure Machine Learning Pipelines](https://docs.microsoft.com/en-us/azure/machine-learning/service/concept-ml-pipelines) overview, or the [readme article](../README.md).\n",
"To get more information about Azure machine learning pipelines, please read our [Azure Machine Learning Pipelines](https://aka.ms/pl-concept) overview, or the [readme article](https://aka.ms/pl-readme).\n",
"\n",
"In this notebook, we provide a gentle introduction to Azure machine learning pipelines. We build a pipeline that runs jobs unattended on different compute clusters; in this notebook, you'll see how to use the basic Azure ML SDK APIs for constructing this pipeline.\n",
" "
@@ -44,7 +44,7 @@
"metadata": {},
"source": [
"## Prerequisites and Azure Machine Learning Basics\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration notebook](../../../configuration.ipynb) located at https://github.com/Azure/MachineLearningNotebooks first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n"
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration notebook](https://aka.ms/pl-config) first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n"
]
},
{
@@ -127,7 +127,7 @@
"metadata": {},
"source": [
"### Required data and script files for the the tutorial\n",
"Sample files required to finish this tutorial are already copied to the corresponding source_directory locations. Even though the .py provided in the samples don't have much \"ML work,\" as a data scientist, you will work on this extensively as part of your work. To complete this tutorial, the contents of these files are not very important. The one-line files are for demostration purpose only."
"Sample files required to finish this tutorial are already copied to the corresponding source_directory locations. Even though the .py provided in the samples does not have much \"ML work\" as a data scientist, you will work on this extensively as part of your work. To complete this tutorial, the contents of these files are not very important. The one-line files are for demostration purpose only."
]
},
{
@@ -597,7 +597,7 @@
"metadata": {},
"source": [
"# Next: Pipelines with data dependency\n",
"The next [notebook](./aml-pipelines-with-data-dependency-steps.ipynb) demostrates how to construct a pipeline with data dependency."
"The next [notebook](https://aka.ms/pl-data-dep) demostrates how to construct a pipeline with data dependency."
]
}
],

2
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-how-to-use-azurebatch-to-run-a-windows-executable.ipynb
View File

@@ -74,7 +74,7 @@
"source": [
"Initialize a workspace object from persisted configuration. Make sure the config file is present at .\\config.json\n",
"\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, If you don't have a config.json file, please go through the configuration Notebook located [here](https://github.com/Azure/MachineLearningNotebooks). \n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, If you don't have a config.json file, please go through the [configuration Notebook](https://aka.ms/pl-config) located [here](https://github.com/Azure/MachineLearningNotebooks). \n",
"\n",
"This sets you up with a working config file that has information on your workspace, subscription id, etc. "
]

2
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-how-to-use-estimatorstep.ipynb
View File

@@ -27,7 +27,7 @@
"\n",
"## Prerequisite:\n",
"* Understand the [architecture and terms](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture) introduced by Azure Machine Learning\n",
"* If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, go through the [configuration notebook](../../../configuration.ipynb) to:\n",
"* If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, go through the [configuration notebook](https://aka.ms/pl-config) to:\n",
" * install the AML SDK\n",
" * create a workspace and its configuration file (`config.json`)"
]

2
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-parameter-tuning-with-hyperdrive.ipynb
View File

@@ -30,7 +30,7 @@
"metadata": {},
"source": [
"## Prerequisites and Azure Machine Learning Basics\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the configuration Notebook located at https://github.com/Azure/MachineLearningNotebooks first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration Notebook](https://aka.ms/pl-config) first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n",
"\n",
"## Azure Machine Learning and Pipeline SDK-specific imports"
]

4
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-publish-and-run-using-rest-endpoint.ipynb
View File

@@ -28,7 +28,7 @@
"metadata": {},
"source": [
"## Prerequisites and Azure Machine Learning Basics\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the configuration Notebook located at https://github.com/Azure/MachineLearningNotebooks first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration Notebook](https://aka.ms/pl-config) first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n",
"\n",
"### Initialization Steps"
]
@@ -393,7 +393,7 @@
"metadata": {},
"source": [
"# Next: Data Transfer\n",
"The next [notebook](./aml-pipelines-data-transfer.ipynb) will showcase data transfer steps between different types of data stores."
"The next [notebook](https://aka.ms/pl-data-trans) will showcase data transfer steps between different types of data stores."
]
}
],

2
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-setup-schedule-for-a-published-pipeline.ipynb
View File

@@ -28,7 +28,7 @@
"metadata": {},
"source": [
"## Prerequisites and AML Basics\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the configuration Notebook located at https://github.com/Azure/MachineLearningNotebooks first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc.\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration Notebook](https://aka.ms/pl-config) first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc.\n",
"\n",
"### Initialization Steps"
]

2
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-setup-versioned-pipeline-endpoints.ipynb
View File

@@ -32,7 +32,7 @@
"metadata": {},
"source": [
"### Prerequisites and AML Basics\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration Notebook](https://github.com/Azure/MachineLearningNotebooks) first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc.\n"
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration Notebook](https://aka.ms/pl-config) first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc.\n"
]
},
{

39
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-use-databricks-as-compute-target.ipynb
View File

@@ -20,7 +20,7 @@
"metadata": {},
"source": [
"# Using Databricks as a Compute Target from Azure Machine Learning Pipeline\n",
"To use Databricks as a compute target from [Azure Machine Learning Pipeline](https://docs.microsoft.com/en-us/azure/machine-learning/service/concept-ml-pipelines), a [DatabricksStep](https://docs.microsoft.com/en-us/python/api/azureml-pipeline-steps/azureml.pipeline.steps.databricks_step.databricksstep?view=azure-ml-py) is used. This notebook demonstrates the use of DatabricksStep in Azure Machine Learning Pipeline.\n",
"To use Databricks as a compute target from [Azure Machine Learning Pipeline](https://aka.ms/pl-concept), a [DatabricksStep](https://docs.microsoft.com/en-us/python/api/azureml-pipeline-steps/azureml.pipeline.steps.databricks_step.databricksstep?view=azure-ml-py) is used. This notebook demonstrates the use of DatabricksStep in Azure Machine Learning Pipeline.\n",
"\n",
"The notebook will show:\n",
"1. Running an arbitrary Databricks notebook that the customer has in Databricks workspace\n",
@@ -290,7 +290,7 @@
"metadata": {},
"source": [
"## Use Databricks from Azure Machine Learning Pipeline\n",
"To use Databricks as a compute target from Azure Machine Learning Pipeline, a DatabricksStep is used. Let's define a datasource (via DataReference) and intermediate data (via PipelineData) to be used in DatabricksStep."
"To use Databricks as a compute target from Azure Machine Learning Pipeline, a DatabricksStep is used. Let's define a datasource (via DataReference), intermediate data (via PipelineData) and a pipeline parameter (via PipelineParameter) to be used in DatabricksStep."
]
},
{
@@ -299,10 +299,14 @@
"metadata": {},
"outputs": [],
"source": [
"from azureml.pipelince.core import PipelineParameter\n",
"\n",
"# Use the default blob storage\n",
"def_blob_store = Datastore(ws, \"workspaceblobstore\")\n",
"print('Datastore {} will be used'.format(def_blob_store.name))\n",
"\n",
"pipeline_param = PipelineParameter(name=\"my_pipeline_param\", default_value=\"pipeline_param1\")\n",
"\n",
"# We are uploading a sample file in the local directory to be used as a datasource\n",
"def_blob_store.upload_files(files=[\"./testdata.txt\"], target_path=\"dbtest\", overwrite=False)\n",
"\n",
@@ -406,7 +410,9 @@
"### 1. Running the demo notebook already added to the Databricks workspace\n",
"Create a notebook in the Azure Databricks workspace, and provide the path to that notebook as the value associated with the environment variable \"DATABRICKS_NOTEBOOK_PATH\". This will then set the variable\u00c2\u00a0notebook_path\u00c2\u00a0when you run the code cell below:\n",
"\n",
"your notebook's path in Azure Databricks UI by hovering over to notebook's title. A typical path of notebook looks like this `/Users/example@databricks.com/example`. See [Databricks Workspace](https://docs.azuredatabricks.net/user-guide/workspace.html) to learn about the folder structure."
"your notebook's path in Azure Databricks UI by hovering over to notebook's title. A typical path of notebook looks like this `/Users/example@databricks.com/example`. See [Databricks Workspace](https://docs.azuredatabricks.net/user-guide/workspace.html) to learn about the folder structure.\n",
"\n",
"Note: DataPath `PipelineParameter` should be provided in list of inputs. Such parameters can be accessed by the datapath `name`."
]
},
{
@@ -423,7 +429,8 @@
" outputs=[step_1_output],\n",
" num_workers=1,\n",
" notebook_path=notebook_path,\n",
" notebook_params={'myparam': 'testparam'},\n",
" notebook_params={'myparam': 'testparam', \n",
" 'myparam2': pipeline_param},\n",
" run_name='DB_Notebook_demo',\n",
" compute_target=databricks_compute,\n",
" allow_reuse=True\n",
@@ -434,7 +441,9 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Build and submit the Experiment"
"#### Build and submit the Experiment\n",
"\n",
"Note: Default value of `pipeline_param` will be used if different value is not specified in pipeline parameters during submission"
]
},
{
@@ -479,7 +488,9 @@
"dbfs cp ./train-db-dbfs.py dbfs:/train-db-dbfs.py\n",
"```\n",
"\n",
"The code in the below cell assumes that you have completed the previous step of uploading the script `train-db-dbfs.py` to the root folder in DBFS."
"The code in the below cell assumes that you have completed the previous step of uploading the script `train-db-dbfs.py` to the root folder in DBFS.\n",
"\n",
"Note: `pipeline_param` will add two values in the python_script_params, a name followed by value. the name will be in this format `--MY_PIPELINE_PARAM`. For example, in the given case, python_script_params will be `[\"arg1\", \"--MY_PIPELINE_PARAM\", \"pipeline_param1\", \"arg2\"]`"
]
},
{
@@ -495,7 +506,7 @@
" inputs=[step_1_input],\n",
" num_workers=1,\n",
" python_script_path=python_script_path,\n",
" python_script_params={'--input_data'},\n",
" python_script_params={'arg1', pipeline_param, 'arg2},\n",
" run_name='DB_Python_demo',\n",
" compute_target=databricks_compute,\n",
" allow_reuse=True\n",
@@ -545,7 +556,9 @@
"### 3. Running a Python script in Databricks that currenlty is in local computer\n",
"To run a Python script that is currently in your local computer, follow the instructions below. \n",
"\n",
"The commented out code below code assumes that you have `train-db-local.py` in the `scripts` subdirectory under the current working directory.\n",
"The commented out code below code assumes that you have `train-db-local.py` in the `source_directory` subdirectory under the current working directory. \n",
"\n",
"The best practice is to use separate folders for scripts and its dependent files for each step and specify that folder as the `source_directory` for the step. This helps reduce the size of the snapshot created for the step (only the specific folder is snapshotted). Since changes in any files in the `source_directory` would trigger a re-upload of the snapshot, this helps keep the reuse of the step when there are no changes in the `source_directory` of the step.\n",
"\n",
"In this case, the Python script will be uploaded first to DBFS, and then the script will be run in Databricks."
]
@@ -557,7 +570,7 @@
"outputs": [],
"source": [
"python_script_name = \"train-db-local.py\"\n",
"source_directory = \".\"\n",
"source_directory = \"./databricks_train\"\n",
"\n",
"dbPythonInLocalMachineStep = DatabricksStep(\n",
" name=\"DBPythonInLocalMachine\",\n",
@@ -618,7 +631,9 @@
"\n",
"```\n",
"dbfs cp ./train-db-dbfs.jar dbfs:/train-db-dbfs.jar\n",
"```"
"```\n",
"\n",
"Note: `pipeline_param` will add two values in the python_script_params, a name followed by value. the name will be in this format `--MY_PIPELINE_PARAM`. For example, in the given case, python_script_params will be `[\"arg1\", \"--MY_PIPELINE_PARAM\", \"pipeline_param1\", \"arg2\"]`"
]
},
{
@@ -635,7 +650,7 @@
" inputs=[step_1_input],\n",
" num_workers=1,\n",
" main_class_name=main_jar_class_name,\n",
" jar_params={'arg1', 'arg2'},\n",
" jar_params={'arg1', pipeline_param, 'arg2'},\n",
" run_name='DB_JAR_demo',\n",
" jar_libraries=[JarLibrary(jar_library_dbfs_path)],\n",
" compute_target=databricks_compute,\n",
@@ -684,7 +699,7 @@
"metadata": {},
"source": [
"# Next: ADLA as a Compute Target\n",
"To use ADLA as a compute target from Azure Machine Learning Pipeline, a AdlaStep is used. This [notebook](./aml-pipelines-use-adla-as-compute-target.ipynb) demonstrates the use of AdlaStep in Azure Machine Learning Pipeline."
"To use ADLA as a compute target from Azure Machine Learning Pipeline, a AdlaStep is used. This [notebook](https://aka.ms/pl-adla) demonstrates the use of AdlaStep in Azure Machine Learning Pipeline."
]
}
],

205
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-with-automated-machine-learning-step.ipynb
View File

@@ -28,24 +28,19 @@
"metadata": {},
"source": [
"## Introduction\n",
"In this example we use the scikit-learn's [digit dataset](http://scikit-learn.org/stable/datasets/index.html#optical-recognition-of-handwritten-digits-dataset) to showcase how you can use AutoML for a simple classification problem.\n",
"In this example we showcase how you can use the `azureml.dataprep` SDK to load and prepare data for AutoML via AML Pipeline. `azureml.dataprep` can also be used standalone; full documentation can be found [here](https://github.com/Microsoft/PendletonDocs).\n",
"\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you have executed the [configuration](../../../configuration.ipynb) before running this notebook.\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you have executed the [configuration](https://aka.ms/pl-config) before running this notebook.\n",
"\n",
"In this notebook you would see\n",
"In this notebook you will learn how to:\n",
"1. Create an `Experiment` in an existing `Workspace`.\n",
"2. Create or Attach existing AmlCompute to a workspace.\n",
"3. Configure AutoML using `AutoMLConfig`.\n",
"4. Use AutoMLStep\n",
"5. Train the model using AmlCompute\n",
"6. Explore the results.\n",
"7. Test the best fitted model.\n",
"\n",
"In addition this notebook showcases the following features\n",
"- **Parallel** executions for iterations\n",
"- **Asynchronous** tracking of progress\n",
"- Retrieving models for any iteration or logged metric\n",
"- Specifying AutoML settings as `**kwargs`"
"3. Define data loading and preparation steps in a `Dataflow` using `azureml.dataprep`.\n",
"4. Configure AutoML using `AutoMLConfig`.\n",
"5. Use AutoMLStep\n",
"6. Train the model using AmlCompute\n",
"7. Explore the results.\n",
"8. Test the best fitted model."
]
},
{
@@ -69,6 +64,8 @@
"import numpy as np\n",
"import pandas as pd\n",
"from sklearn import datasets\n",
"import pkg_resources\n",
"import azureml.dataprep as dprep\n",
"\n",
"import azureml.core\n",
"from azureml.core.experiment import Experiment\n",
@@ -131,7 +128,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create or Attach existing AmlCompute\n",
"### Create or Attach an AmlCompute cluster\n",
"You will need to create a [compute target](https://docs.microsoft.com/azure/machine-learning/service/concept-azure-machine-learning-architecture#compute-target) for your AutoML run. In this tutorial, you get the default `AmlCompute` as your training compute resource."
]
},
@@ -168,45 +165,6 @@
" # For a more detailed view of current AmlCompute status, use get_status()."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prepare and Point to Data\n",
"For remote executions, you need to make the data accessible from the remote compute.\n",
"This can be done by uploading the data to DataStore.\n",
"In this example, we upload scikit-learn's [load_digits](http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_digits.html) data."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"data_train = datasets.load_digits()\n",
"\n",
"if not os.path.isdir('data'):\n",
" os.mkdir('data')\n",
" \n",
"if not os.path.exists(project_folder):\n",
" os.makedirs(project_folder)\n",
" \n",
"pd.DataFrame(data_train.data).to_csv(\"data/X_train.tsv\", index=False, header=False, quoting=csv.QUOTE_ALL, sep=\"\\t\")\n",
"pd.DataFrame(data_train.target).to_csv(\"data/y_train.tsv\", index=False, header=False, sep=\"\\t\")\n",
"\n",
"ds = ws.get_default_datastore()\n",
"ds.upload(src_dir='./data', target_path='bai_data', overwrite=True, show_progress=True)\n",
"\n",
"from azureml.data.data_reference import DataReference \n",
"input_data = DataReference(datastore=ds, \n",
" data_reference_name=\"input_data_reference\",\n",
" path_on_datastore='bai_data',\n",
" mode='download',\n",
" path_on_compute='/tmp/azureml_runs',\n",
" overwrite=False)"
]
},
{
"cell_type": "code",
"execution_count": null,
@@ -216,54 +174,77 @@
"# create a new RunConfig object\n",
"conda_run_config = RunConfiguration(framework=\"python\")\n",
"\n",
"# Set compute target to AmlCompute\n",
"#conda_run_config.target = compute_target\n",
"\n",
"conda_run_config.environment.docker.enabled = True\n",
"conda_run_config.environment.docker.base_image = azureml.core.runconfig.DEFAULT_CPU_IMAGE\n",
"\n",
"cd = CondaDependencies.create(pip_packages=['azureml-sdk[automl]'], \n",
" conda_packages=['numpy', 'py-xgboost'], \n",
" pin_sdk_version=False)\n",
" conda_packages=['numpy', 'py-xgboost<=0.80'])\n",
"conda_run_config.environment.python.conda_dependencies = cd\n",
"\n",
"print('run config is ready')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Data"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%%writefile $project_folder/get_data.py\n",
"\n",
"import pandas as pd\n",
"\n",
"def get_data():\n",
" X_train = pd.read_csv(\"/tmp/azureml_runs/bai_data/X_train.tsv\", delimiter=\"\\t\", header=None, quotechar='\"')\n",
" y_train = pd.read_csv(\"/tmp/azureml_runs/bai_data/y_train.tsv\", delimiter=\"\\t\", header=None, quotechar='\"')\n",
"\n",
" return { \"X\" : X_train.values, \"y\" : y_train.values.flatten() }\n"
"# You can use `auto_read_file` which intelligently figures out delimiters and datatypes of a file.\n",
"# The data referenced here was a 1MB simple random sample of the Chicago Crime data into a local temporary directory.\n",
"# You can also use `read_csv` and `to_*` transformations to read (with overridable delimiter)\n",
"# and convert column types manually.\n",
"example_data = 'https://dprepdata.blob.core.windows.net/demo/crime0-random.csv'\n",
"dflow = dprep.auto_read_file(example_data).skip(1) # Remove the header row.\n",
"dflow.get_profile()"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# As `Primary Type` is our y data, we need to drop the values those are null in this column.\n",
"dflow = dflow.drop_nulls('Primary Type')\n",
"dflow.head(5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Set up AutoMLConfig for Training\n",
"### Review the Data Preparation Result\n",
"\n",
"You can specify `automl_settings` as `**kwargs` as well. Also note that you can use a `get_data()` function for local excutions too.\n",
"You can peek the result of a Dataflow at any range using `skip(i)` and `head(j)`. Doing so evaluates only `j` records for all the steps in the Dataflow, which makes it fast even against large datasets.\n",
"\n",
"**Note:** When using AmlCompute, you can't pass Numpy arrays directly to the fit method.\n",
"\n",
"|Property|Description|\n",
"|-|-|\n",
"|**primary_metric**|This is the metric that you want to optimize. Classification supports the following primary metrics: <br><i>accuracy</i><br><i>AUC_weighted</i><br><i>average_precision_score_weighted</i><br><i>norm_macro_recall</i><br><i>precision_score_weighted</i>|\n",
"|**iteration_timeout_minutes**|Time limit in minutes for each iteration.|\n",
"|**iterations**|Number of iterations. In each iteration AutoML trains a specific pipeline with the data.|\n",
"|**n_cross_validations**|Number of cross validation splits.|\n",
"|**max_concurrent_iterations**|Maximum number of iterations that would be executed in parallel. This should be less than the number of cores on the DSVM.|"
"`Dataflow` objects are immutable and are composed of a list of data preparation steps. A `Dataflow` object can be branched at any point for further usage."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"X = dflow.drop_columns(columns=['Primary Type', 'FBI Code'])\n",
"y = dflow.keep_columns(columns=['Primary Type'], validate_column_exists=True)\n",
"print('X and y are ready!')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Train\n",
"This creates a general AutoML settings object."
]
},
{
@@ -274,11 +255,9 @@
"source": [
"automl_settings = {\n",
" \"iteration_timeout_minutes\" : 5,\n",
" \"iterations\": 20,\n",
" \"n_cross_validations\": 5,\n",
" \"iterations\" : 2,\n",
" \"primary_metric\" : 'AUC_weighted',\n",
" \"preprocess\": False,\n",
" \"max_concurrent_iterations\": 3,\n",
" \"preprocess\" : True,\n",
" \"verbosity\" : logging.INFO\n",
"}\n",
"automl_config = AutoMLConfig(task = 'classification',\n",
@@ -286,7 +265,8 @@
" path = project_folder,\n",
" compute_target=compute_target,\n",
" run_configuration=conda_run_config,\n",
" data_script = project_folder + \"/get_data.py\",\n",
" X = X,\n",
" y = y,\n",
" **automl_settings\n",
" )"
]
@@ -295,15 +275,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"Call the `submit` method on the experiment object and pass the run configuration. For remote runs the execution is asynchronous, so you will see the iterations get populated as they complete. You can interact with the widgets and models even when the experiment is running to retrieve the best model up to that point. Once you are satisfied with the model, you can cancel a particular iteration or the whole run.\n",
"In this example, we specify `show_output = False` to suppress console output while the run is in progress."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Define AutoMLStep"
"You can define outputs for the AutoMLStep using TrainingOutput."
]
},
{
@@ -314,6 +286,7 @@
"source": [
"from azureml.pipeline.core import PipelineData, TrainingOutput\n",
"\n",
"ds = ws.get_default_datastore()\n",
"metrics_output_name = 'metrics_output'\n",
"best_model_output_name = 'best_model_output'\n",
"\n",
@@ -327,6 +300,13 @@
" training_output=TrainingOutput(type='Model'))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Create an AutoMLStep."
]
},
{
"cell_type": "code",
"execution_count": null,
@@ -335,9 +315,7 @@
"source": [
"automl_step = AutoMLStep(\n",
" name='automl_module',\n",
" experiment=experiment,\n",
" automl_config=automl_config,\n",
" inputs=[input_data],\n",
" outputs=[metirics_data, model_data],\n",
" allow_reuse=True)"
]
@@ -453,7 +431,8 @@
"metadata": {},
"source": [
"### Test the Model\n",
"#### Load Test Data"
"#### Load Test Data\n",
"For the test data, it should have the same preparation step as the train data. Otherwise it might get failed at the preprocessing step."
]
},
{
@@ -462,17 +441,17 @@
"metadata": {},
"outputs": [],
"source": [
"digits = datasets.load_digits()\n",
"X_test = digits.data[:10, :]\n",
"y_test = digits.target[:10]\n",
"images = digits.images[:10]"
"dflow_test = dprep.auto_read_file(path='https://dprepdata.blob.core.windows.net/demo/crime0-test.csv').skip(1)\n",
"dflow_test = dflow_test.drop_nulls('Primary Type')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Testing Best Model"
"#### Testing Our Best Fitted Model\n",
"\n",
"We will use confusion matrix to see how our model works."
]
},
{
@@ -481,17 +460,19 @@
"metadata": {},
"outputs": [],
"source": [
"# Randomly select digits and test.\n",
"for index in np.random.choice(len(y_test), 3, replace = False):\n",
" print(index)\n",
" predicted = best_model.predict(X_test[index:index + 1])[0]\n",
" label = y_test[index]\n",
" title = \"Label value = %d Predicted value = %d \" % (label, predicted)\n",
" fig = plt.figure(1, figsize=(3,3))\n",
" ax1 = fig.add_axes((0,0,.8,.8))\n",
" ax1.set_title(title)\n",
" plt.imshow(images[index], cmap = plt.cm.gray_r, interpolation = 'nearest')\n",
" plt.show()"
"from pandas_ml import ConfusionMatrix\n",
"\n",
"y_test = dflow_test.keep_columns(columns=['Primary Type']).to_pandas_dataframe()\n",
"X_test = dflow_test.drop_columns(columns=['Primary Type', 'FBI Code']).to_pandas_dataframe()\n",
"\n",
"\n",
"ypred = best_model.predict(X_test)\n",
"\n",
"cm = ConfusionMatrix(y_test['Primary Type'], ypred)\n",
"\n",
"print(cm)\n",
"\n",
"cm.plot()"
]
}
],

4
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-with-data-dependency-steps.ipynb
View File

@@ -28,7 +28,7 @@
"metadata": {},
"source": [
"## Prerequisites and Azure Machine Learning Basics\n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the configuration Notebook located at https://github.com/Azure/MachineLearningNotebooks first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n",
"If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, make sure you go through the [configuration Notebook](https://aka.ms/pl-config) first if you haven't. This sets you up with a working config file that has information on your workspace, subscription id, etc. \n",
"\n",
"### Azure Machine Learning and Pipeline SDK-specific Imports"
]
@@ -517,7 +517,7 @@
"metadata": {},
"source": [
"# Next: Publishing the Pipeline and calling it from the REST endpoint\n",
"See this [notebook](./aml-pipelines-publish-and-run-using-rest-endpoint.ipynb) to understand how the pipeline is published and you can call the REST endpoint to run the pipeline."
"See this [notebook](https://aka.ms/pl-pub-rep) to understand how the pipeline is published and you can call the REST endpoint to run the pipeline."
]
}
],

5
how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/databricks_train/train-db-local.py Normal file
View File

@@ -0,0 +1,5 @@
# Copyright (c) Microsoft. All rights reserved.
# Licensed under the MIT license.
print("In train.py")
print("As a data scientist, this is where I use my training code.")

1
how-to-use-azureml/training-with-deep-learning/README.md
View File

@@ -14,6 +14,7 @@ These examples show you:
10. [Distributed training using Chainer](distributed-chainer)
11. [Export run history records to Tensorboard](export-run-history-to-tensorboard)
12. [Use TensorBoard to monitor training execution](tensorboard)
13. [Resuming training from previous run](train-tensorflow-resume-training)
Learn more about how to use `Estimator` class to [train deep neural networks with Azure Machine Learning](https://docs.microsoft.com/azure/machine-learning/service/how-to-train-ml-models).

2
how-to-use-azureml/training/logging-api/logging-api.ipynb
View File

@@ -100,7 +100,7 @@
"\n",
"# Check core SDK version number\n",
"\n",
"print(\"This notebook was created using SDK version 1.0.53, you are currently running version\", azureml.core.VERSION)"
"print(\"This notebook was created using SDK version 1.0.55, you are currently running version\", azureml.core.VERSION)"
]
},
{

0
model-deployment/README.md Normal file
View File

2
setup-environment/configuration.ipynb
View File

@@ -102,7 +102,7 @@
"source": [
"import azureml.core\n",
"\n",
"print(\"This notebook was created using version 1.0.53 of the Azure ML SDK\")\n",
"print(\"This notebook was created using version 1.0.55 of the Azure ML SDK\")\n",
"print(\"You are currently using version\", azureml.core.VERSION, \"of the Azure ML SDK\")"
]
},

0
training/README.md Normal file
View File

5
tutorials/README.md
View File

@@ -8,6 +8,11 @@ two sets of tutorial articles for:
If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, run the [configuration Notebook](../configuration.ipynb) notebook first to set up your Azure ML Workspace. Then, run the notebooks in following recommended order.
### Create first ML experiment
* [Part 1](https://docs.microsoft.com/azure/machine-learning/service/tutorial-quickstart-setup): Set up workspace & dev environment
* [Part 2](tutorial-quickstart-train-model.ipynb): Learn the foundational design patterns in Azure Machine Learning service, and train a simple scikit-learn model based on the diabetes data set
### Image classification
* [Part 1](img-classification-part1-training.ipynb): Train an image classification model with Azure Machine Learning.

16
tutorials/img-classification-part1-training.ipynb
View File

@@ -218,7 +218,7 @@
"source": [
"### Display some sample images\n",
"\n",
"Load the compressed files into `numpy` arrays. Then use `matplotlib` to plot 30 random images from the dataset with their labels above them. Note this step requires a `load_data` function that's included in an `util.py` file. This file is included in the sample folder. Please make sure it is placed in the same folder as this notebook. The `load_data` function simply parses the compresse files into numpy arrays."
"Load the compressed files into `numpy` arrays. Then use `matplotlib` to plot 30 random images from the dataset with their labels above them. Note this step requires a `load_data` function that's included in an `utils.py` file. This file is included in the sample folder. Please make sure it is placed in the same folder as this notebook. The `load_data` function simply parses the compresse files into numpy arrays."
]
},
{
@@ -260,7 +260,7 @@
"\n",
"Now make the data accessible remotely by uploading that data from your local machine into Azure so it can be accessed for remote training. The datastore is a convenient construct associated with your workspace for you to upload/download data, and interact with it from your remote compute targets. It is backed by Azure blob storage account.\n",
"\n",
"The MNIST files are uploaded into a directory named `mnist` at the root of the datastore."
"The MNIST files are uploaded into a directory named `mnist` at the root of the datastore. See [access data from your datastores](https://docs.microsoft.com/bs-latn-ba/azure/machine-learning/service/how-to-access-data) for more information."
]
},
{
@@ -674,6 +674,18 @@
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.6"
},
"msauthor": "roastala"
},
"nbformat": 4,

2
tutorials/tutorial-1st-experiment-sdk-train.ipynb
View File

@@ -11,7 +11,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/tutorials/tutorial-1st-experiment-sdk-train.png)"
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks/tutorials/tutorial-quickstart-train-model.png)"
]
},
{

5
tutorials/tutorial-1st-experiment-sdk-train.yml Normal file
View File

@@ -0,0 +1,5 @@
name: tutorial-1st-experiment-sdk-train
dependencies:
- pip:
- azureml-sdk
- sklearn

45
work-with-data/dataprep/README.md
View File

@@ -31,6 +31,51 @@ If you have any questions or feedback, send us an email at: [askamldataprep@micr
## Release Notes
### 2019-07-25 (version 1.1.9)
New features
- Added support for reading a file directly from a http or https url.
Bug fixes and improvements
- Improved error message when attempting to read a Parquet Dataset from a remote source (which is not currently supported).
- Fixed a bug when writing to Parquet file format in ADLS Gen 2, and updating the ADLS Gen 2 container name in the path.
### 2019-07-09 (version 1.1.8)
New features
- Dataflow objects can now be iterated over, producing a sequence of records. See documentation for `Dataflow.to_record_iterator`.
Bug fixes and improvements
- Increased the robustness of DataPrep SDK.
- Improved handling of pandas DataFrames with non-string Column Indexes.
- Improved the performance of `to_pandas_dataframe` in Datasets.
- Fixed a bug where Spark execution of Datasets failed when run in a multi-node environment.
### 2019-07-01 (version 1.1.7)
We reverted a change that improved performance, as it was causing issues for some customers using Azure Databricks. If you experienced an issue on Azure Databricks, you can upgrade to version 1.1.7 using one of the methods below:
1. Run this script to upgrade: `%sh /home/ubuntu/databricks/python/bin/pip install azureml-dataprep==1.1.7`
2. Recreate the cluster, which will install the latest Data Prep SDK version.
### 2019-06-24 (version 1.1.6)
New features
- Added summary functions for top values (`SummaryFunction.TOPVALUES`) and bottom values (`SummaryFunction.BOTTOMVALUES`).
Bug fixes and improvements
- Significantly improved the performance of `read_pandas_dataframe`.
- Fixed a bug that would cause `get_profile()` on a Dataflow pointing to binary files to fail.
- Exposed `set_diagnostics_collection()` to allow for programmatic enabling/disabling of the telemetry collection.
- Changed the behavior of `get_profile()`. NaN values are now ignored for Min, Mean, Std, and Sum, which aligns with the behavior of Pandas.
### 2019-06-10 (version 1.1.5)
Bug fixes and improvements
- For interpreted datetime values that have a 2-digit year format, the range of valid years has been updated to match Windows May Release. The range has been changed from 1930-2029 to 1950-2049.
- When reading in a file and setting `handleQuotedLineBreaks=True`, `\r` will be treated as a new line.
- Fixed a bug that caused `read_pandas_dataframe` to fail in some cases.
- Improved performance of `get_profile`.
- Improved error messages.
### 2019-05-28 (version 1.1.4)
New features

34
work-with-data/dataprep/how-to-guides/data-ingestion.ipynb
View File

@@ -48,7 +48,8 @@
"[Read From Azure Blob](#azure-blob)<br>\n",
"[Read From ADLS](#adls)<br>\n",
"[Read From ADLSGen2](#adlsgen2)<br>\n",
"[Read Pandas DataFrame](#pandas-df)<br>"
"[Read Pandas DataFrame](#pandas-df)<br>\n",
"[Read From HTTP/HTTPS Link](#http)<br>"
]
},
{
@@ -1047,6 +1048,37 @@
"source": [
"dflow_df.head(5)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"<a id=\"http\"></a>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Read from HTTP/HTTPS Link"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can pass in an HTTP/HTTPS path when loading remote data source."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"dflow = dprep.read_csv('https://dprepdata.blob.core.windows.net/test/Sample-Spreadsheet-10-rows.csv')\n",
"dflow.head(5)"
]
}
],
"metadata": {

796
work-with-data/datasets/datasets-diff/datasets-diff.ipynb Normal file
View File

@@ -0,0 +1,796 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Copyright (c) Microsoft Corporation. All rights reserved.\n",
"\n",
"Licensed under the MIT License."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![Impressions](https://PixelServer20190423114238.azurewebsites.net/api/impressions/MachineLearningNotebooks//notebooks/work-with-data/datasets/datasets-tutorial/datasets-diff.png)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# <center>Detect drift using Dataset Diff API </center>\n",
"\n",
"<br>\n",
"<font size=2>\n",
" This notebook provides step by step instructions on how to compare two different datasets. It includes two parts\u00ef\u00bc\u0161\n",
" <br>&nbsp;&nbsp;&nbsp;&nbsp;&#x2611; compare two datasets using local compute;\n",
" <br>&nbsp;&nbsp;&nbsp;&nbsp;&#x2611; compare two datasets remotely using Azure ML compute.\n",
"</font>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Prerequisites and Setup\n",
"\n",
"<font size=2>This section is shared by both local and remote execution, you may need duplicate this section if splitting this notebook into separate local/remote notebooks.</font>\n",
"\n",
"\n",
"## Prerequisites\n",
"\n",
"### Install Supporting Packages"
]
},
{
"cell_type": "markdown",
"metadata": {
"scrolled": true
},
"source": [
"&nbsp;&nbsp;&nbsp;&nbsp;pip install scipy<br>\n",
"&nbsp;&nbsp;&nbsp;&nbsp;pip install tqdm<br>\n",
"&nbsp;&nbsp;&nbsp;&nbsp;pip install pandas<br>\n",
"&nbsp;&nbsp;&nbsp;&nbsp;pip install pyarrow<br>\n",
"&nbsp;&nbsp;&nbsp;&nbsp;pip install ipywidgets<br>\n",
"&nbsp;&nbsp;&nbsp;&nbsp;pip install lightgbm<br>\n",
"&nbsp;&nbsp;&nbsp;&nbsp;pip install matplotlib<br>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Install AzureML Packages"
]
},
{
"cell_type": "markdown",
"metadata": {
"scrolled": true
},
"source": [
"&nbsp;&nbsp;&nbsp;&nbsp;pip install --user azureml-core<br>\n",
"\n",
"&nbsp;&nbsp;&nbsp;&nbsp;pip install --user azureml-opendatasets<br>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Import Dependencies"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"import sys\n",
"import warnings\n",
"import requests\n",
"import pandas as pd\n",
"import numpy as np\n",
"import ipywidgets as widgets\n",
"\n",
"import azureml.core\n",
"\n",
"from io import StringIO\n",
"from tqdm import tqdm\n",
"from IPython import display\n",
"from datetime import datetime, timedelta\n",
"from azureml.core import Datastore, Dataset\n",
"from azureml.opendatasets import NoaaIsdWeather\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Declare Variables For Demo\n",
"\n",
"Feel free to customize them."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"year = 2016\n",
"month = 1\n",
"date = 1\n",
"b_days = 2 # for baseline\n",
"t_days = 7 # for target\n",
"\n",
"local_folder = \"demo\"\n",
"baseline_file = 'baseline.csv'\n",
"\n",
"feature_columns = ['usaf', 'wban', 'latitude', 'longitude', 'elevation', 'temperature', 'p_k']"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prepare Datasets\n",
"\n",
"<font size=2>The diff calcualtion is always between two datasets, here for demo, we use \"baseline\" and \"target\" to present them.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"os.makedirs(local_folder, exist_ok=True)\n",
"\n",
"local_baseline = os.path.join(local_folder, baseline_file)\n",
"\n",
"start_date = datetime(year, month, date)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Prepare Baseline Dataset\n",
"<font size=2>Retrieve wether data from NOAA for declared days (b_days declared in above cell). It may takes 2 minutes for 2 days.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"start = start_date\n",
"isd = NoaaIsdWeather(start, start + timedelta(days=b_days))\n",
"\n",
"baseline_df = isd.to_pandas_dataframe()\n",
"baseline_df.head()\n",
"\n",
"baseline_df.to_csv(local_baseline)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Prepare Target Dataset(s)\n",
"\n",
"<font size=2>Retrieve wether data from NOAA for declared days (t_days declared in above cell). It may takes 5 minutes for 7 days.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"for day in tqdm(range(0, t_days)):\n",
" start = start_date + timedelta(days=day)\n",
" isd = NoaaIsdWeather(start, start + timedelta(days=1))\n",
"\n",
" target_df = isd.to_pandas_dataframe()\n",
" target_df = target_df[feature_columns]\n",
" target_df.to_csv(os.path.join(local_folder, 'target_{}.csv'.format(day)))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Predefine Methods For Result Processing\n",
"\n",
"## Parse and Present Datasets' Diff Results\n",
"\n",
"<font size=2>Each diff result is a list of \"DiffMetric\" objects. Typically each objec present a detailed measurement output for a specific column.\n",
"<br><br>Below is an example of \"DiffMetric\" object:</font>\n",
"<font face=\"monospace\" size=1>\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;{&nbsp;&nbsp;\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'name':'percentage_difference_median',&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;measurement&nbsp;name\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'value':0.01270670472603889,&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;the result value a number to indicate how big the diff is for current measurement.\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'extended_properties':{&nbsp;&nbsp;\n",
"</font><font face=\"monospace\" size=1 color=LightSteelBlue><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'action_id':'3d3da05d-0871-4cc9-93cb-f43859aae13b',&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;(remote&nbsp;calculation&nbsp;only)&nbsp;action&nbsp;id\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'from_dataset_id':'12edc566-8803-4e0f-ba91-c2ee05eeddee',&nbsp;&nbsp;-->&nbsp;(remote&nbsp;calculation&nbsp;only)&nbsp;baseline&nbsp;dataset\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'from_dataset_version':'1',&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;(remote&nbsp;calculation&nbsp;only)&nbsp;baseline&nbsp;version\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'to_dataset_id':'9b85c9ba-50c2-4227-a9bc-91dee4a18228',&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;(remote&nbsp;calculation&nbsp;only)&nbsp;target&nbsp;dataset\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'to_dataset_version':'1',&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;(remote&nbsp;calculation&nbsp;only)&nbsp;target&nbsp;version\n",
"</font><font face=\"monospace\" size=1><br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'column_name':'elevation',&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;column&nbsp;name&nbsp;in&nbsp;dataset,&nbsp;<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;could&nbsp;be&nbsp;['name':'datadrift_coefficient']&nbsp;for&nbsp;dataset&nbsp;level&nbsp;diff\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;'metric_category':'profile_diff'&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;-->&nbsp;category,&nbsp;could&nbsp;be&nbsp;:<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;dataset_drift (dataset level)<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;profile_diff (column level)<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;statistical_distance (column level)\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;}\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;}\n",
"</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def parse_result(rst, columns, measurements):\n",
" columnlist = list(columns)\n",
" columnlist.insert(0, \"measurements \\ columns\")\n",
" measurementlist = list(measurements)\n",
" \n",
" daily_result = []\n",
" daily_result.append(columnlist)\n",
" \n",
" drift = None\n",
" daily_contribution = {}\n",
" \n",
" for m in measurements:\n",
" emptylist = ([''] * len(columns))\n",
" emptylist.insert(0, m)\n",
" daily_result.append(emptylist)\n",
"\n",
" for r in rst:\n",
" # get dataset level diff (drift)\n",
" if r.name == \"datadrift_coefficient\":\n",
" drift = r.value\n",
" # get diff (drift) contribution for each column:\n",
" elif r.name == \"datadrift_contribution\":\n",
" daily_contribution[r.extended_properties[\"column_name\"]] = r.value\n",
" # get column level diff measurements\n",
" else:\n",
" if \"column_name\" in r.extended_properties:\n",
" col = r.extended_properties[\"column_name\"]\n",
" msm = r.name\n",
" val = r.value\n",
" cid = columnlist.index(col)\n",
" kid = measurementlist.index(msm) + 1\n",
" daily_result[kid][cid] = val\n",
"\n",
" return daily_result, drift, daily_contribution"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Present Dataset Level Diff (aka drift)\n",
"\n",
"<font size=2>This method will generate two graphs, the left graph presents dataset level difference for all compared baseline-target pairs, the right graph presents dataset level difference contribution for each column so that we know which column impacts more.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"\n",
"import matplotlib.dates as mdates\n",
"import matplotlib.pyplot as plt \n",
"import matplotlib as mpl\n",
"\n",
"def show_diff(drift_metrics, dates, columns, drift_contributions, summary_contribute, bottoms_contribute):\n",
" drifts = [drift_metrics[day] for day in drift_metrics]\n",
" daily_summary_contribution = list(summary_contribute.values())\n",
" xrange = pd.date_range(dates[0], dates[-1], freq='D')\n",
"\n",
" figure = plt.figure(figsize=(16, 4))\n",
" plt.tight_layout()\n",
"\n",
" # left graph\n",
" ax1 = plt.subplot(1, 2, 1)\n",
" ax1.grid()\n",
" plt.sca(ax1)\n",
" plt.title(\"Diff(Drift) Trend\\n\", fontsize=20)\n",
" plt.xticks(rotation=30)\n",
" plt.xlabel(\"Date\", fontsize=16)\n",
" plt.ylabel(\"Drift Coefficent\", fontsize=16)\n",
" plt.plot_date(dates, drifts, '-r', marker='.', linewidth=0.5, markersize=5)\n",
"\n",
" # right graph\n",
" ax2 = plt.subplot(1, 2, 2)\n",
" plt.sca(ax2)\n",
" plt.title(\"Drift Contribution of columns\\n\", fontsize=20)\n",
" plt.xticks(xrange, rotation=30)\n",
" plt.xlabel(\"Date\", fontsize=16)\n",
" plt.ylabel(\"Drift Contribution\", fontsize=16)\n",
"\n",
" yvals = ax2.get_yticks()\n",
" ax2.set_yticklabels(['{:,.2%}'.format(v) for v in yvals])\n",
" ax2.xaxis.set_major_formatter(mdates.DateFormatter('%Y%m-%d'))\n",
"\n",
" for c in columns:\n",
" contribution = []\n",
" for dt in drift_contributions:\n",
" contribution.append(drift_contributions[dt][c])\n",
" bar_ratio = [x / y for x, y in zip(contribution, daily_summary_contribution)]\n",
"\n",
" ax2.bar(dates, height=bar_ratio, bottom=bottoms_contribute)\n",
" bottoms_contribute = [x + y for x, y in zip(bottoms_contribute, bar_ratio)]\n",
"\n",
" plt.legend(columns)\n",
"\n",
" plt.show()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Execute Datasets' Diff Calculation Locally\n",
"\n",
"<font size=2>Local execution let you to run in a Jupyter Notebook or Code editor in a local computer.</font>\n",
"\n",
"## Calculate Dataset Diff At Local\n",
"\n",
"### Create Baseline Dataset\n",
"\n",
"<font size=2>Create baseline dataset object from the retrieved baseline data.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Dataset\n",
"\n",
"baseline = Dataset.auto_read_files(local_baseline, include_path=True)\n",
"\n",
"# The baseline data is not filtered by feature columns list, thus all retrieved data columns will be listed below.\n",
"# You'll see \"Column1\" in the output, which is a default name added when the original column is not available.\n",
"baseline.get_profile()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create Target Datasets\n",
"\n",
"<font size=2>Create target dataset objects from retrieved target data.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"targets = {}\n",
"\n",
"for day in tqdm(range(0, t_days)):\n",
" target = Dataset.auto_read_files(os.path.join(local_folder, 'target_{}.csv'.format(day)))\n",
" targets[day] = target"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Calculate Diff Between Each Target Dataset And Baseline Dataset\n",
"\n",
"<font size=2>Compare each target dataset with baseline dataset to calculate diff between them.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"buf = {}\n",
"\n",
"columns = set()\n",
"measurements = set()\n",
"\n",
"for day in tqdm(range(0, t_days)):\n",
" diff_action = baseline.diff(rhs_dataset=targets[day])\n",
" diff_action.wait_for_completion()\n",
" \n",
" dt = (start_date + timedelta(days=day)).strftime(\"%Y-%m-%d\")\n",
" buf[dt] = diff_action._result\n",
" \n",
" for r in diff_action._result:\n",
" if r.name not in measurements:\n",
" measurements.add(r.name)\n",
" if \"column_name\" in r.extended_properties and r.extended_properties[\"column_name\"] not in columns:\n",
" columns.add(r.extended_properties[\"column_name\"])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Parse And Present Local Execution Results\n",
"\n",
"<font size=2>\n",
"<br>The diff outputs usually contains two different level information:\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;1. General diff, aka dataset level diff. The output is a number between 0 and 1 to indicate what level the diff is. This dataset level diff is also called drift between two datasets.\n",
"<br>&nbsp;&nbsp;&nbsp;&nbsp;2. Detailed diff, aka column level diff. The output is a metrics organized like a 2-D array. One dimension is column names, that is why it's in column level. The other dimension are measurements. The diff calculation actually includes variuos measurements from different perspectives, each measurement will generate an index for each column to present how big impacts this column contributed.\n",
"</font>\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Parse and List Column Level Diff Results\n",
"\n",
"<font size=2>Here will iteratively list all details per each measurement per column calculated.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from pandas import DataFrame\n",
"\n",
"dates = []\n",
"drift_metrics = {}\n",
"drift_contributions = {}\n",
"summary_contribute = {}\n",
"bottoms_contribute = []\n",
"\n",
"for dt, rst in buf.items():\n",
" dates.append(dt)\n",
" print(\"\\n---------------------------------------- Result of {} ----------------------------------------\".format(dt))\n",
" \n",
" daily_result, drift, daily_contribution = parse_result(rst, columns, measurements)\n",
" drift_metrics[dt] = drift\n",
" drift_contributions[dt] = daily_contribution\n",
"\n",
" sum_contribution = 0\n",
" bottoms_contribute.append(0)\n",
" for col, val in daily_contribution.items():\n",
" sum_contribution += val\n",
" summary_contribute[dt] = sum_contribution\n",
"\n",
" \n",
" display.display(pd.DataFrame(daily_result))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Present Dataset Level Diff (aka drift) In Graphs\n",
"\n",
"<font size=2>The left graph presents dataset level difference for all compared baseline-target pairs, the right graph presents dataset level difference contribution for each column so that we know which column impacts more.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"show_diff(drift_metrics, dates, columns, drift_contributions, summary_contribute, bottoms_contribute)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Excute Datasets's Diff Calculation Remotely\n",
"\n",
"<font size=2>Remote execution let you to data compare on more powerful computes - Machine Learning Compute clusters.</font>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Prepare Remote Environment\n",
"### Get Workspace\n",
"<font size=2>\n",
"<br>If you are using an Azure Machine Learning Notebook VM, you are all set. Otherwise, check the <a href=\"../../../configuration.ipynb\" title=\"Create an Azure Machine Learning service workspace\">configuration notebook</a> first if you haven't already to establish your connection to the AzureML Workspace.\n",
"</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.workspace import Workspace\n",
"from azureml.core.authentication import InteractiveLoginAuthentication\n",
"\n",
"ws = Workspace.from_config()\n",
"\n",
"print(ws.name, ws.resource_group, ws.location, ws.subscription_id, sep=\"\\n\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create Compute Resource For Calculation\n",
"<font size=2>Check if compute resouce exists and create a new one if not.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core.compute import AmlCompute, ComputeTarget\n",
"\n",
"existing = False\n",
"del_cmpt = False\n",
"cts = ws.compute_targets\n",
"\n",
"if (ws.DEFAULT_CPU_CLUSTER_NAME in cts and cts[ws.DEFAULT_CPU_CLUSTER_NAME].type == 'AmlCompute'):\n",
" existing = True\n",
" aml_compute = cts[ws.DEFAULT_CPU_CLUSTER_NAME]\n",
" \n",
"if not existing:\n",
" aml_compute = AmlCompute.create(ws,ws.DEFAULT_CPU_CLUSTER_NAME,ws.DEFAULT_CPU_CLUSTER_CONFIGURATION)\n",
" aml_compute.wait_for_completion(show_output=True)\n",
" del_cmpt = True"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Upload Sample Data To Datastore\n",
"\n",
"<font size=2>Upload data files to the blob storage in Azure ML workspace.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Datastore, Dataset\n",
"import azureml.data\n",
"from azureml.data.azure_storage_datastore import AzureFileDatastore, AzureBlobDatastore\n",
"\n",
"remote_data_path ='demo'\n",
"\n",
"dstore = ws.get_default_datastore()\n",
"dstore.upload_files([local_baseline],\n",
" target_path=remote_data_path,\n",
" overwrite=True,\n",
" show_progress=True)\n",
"\n",
"for day in tqdm(range(0, t_days)):\n",
" target_file = os.path.join(local_folder, 'target_{}.csv'.format(day))\n",
" dstore.upload_files([target_file],\n",
" target_path=remote_data_path,\n",
" overwrite=True,\n",
" show_progress=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Register DataSets\n",
"\n",
"<font size=2>Create and Register Datasets.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from azureml.core import Datastore, Dataset\n",
"dstore = ws.get_default_datastore()\n",
"\n",
"xpath = remote_data_path + '/' + baseline_file\n",
"toregister_baseline = Dataset.from_delimited_files(dstore.path(xpath))\n",
"registered_baseline = toregister_baseline.register(workspace = ws,\n",
" name = 'dataset baseline for diff demo',\n",
" description = 'dataset baseline for diff comparison',\n",
" exist_ok = True,\n",
" update_if_exist = True\n",
" )\n",
"\n",
"registered_targets = {}\n",
"for day in tqdm(range(0, t_days)):\n",
" target_file = 'target_{}.csv'.format(day)\n",
" toregister_target = Dataset.from_delimited_files(dstore.path(remote_data_path + '/' + target_file))\n",
" registered_target = toregister_target.register(workspace = ws,\n",
" name = 'dataset target-{} for diff demo'.format(day),\n",
" description = 'target target-{} for diff comparison'.format(day),\n",
" exist_ok = True,\n",
" update_if_exist = True\n",
" )\n",
" registered_targets[day] = registered_target"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Calculate Dataset Diff Remotely\n",
"\n",
"<font size=2>Perform the calculation remotely. This may take 20 minutes.</font>\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"remote_diffs = {}\n",
"\n",
"r_columns = set()\n",
"r_measurements = set()\n",
"\n",
"for day, registered_target in registered_targets.items():\n",
" dt = (start_date + timedelta(days=day)).strftime(\"%Y-%m-%d\")\n",
" remote_diff = registered_baseline.diff(registered_target, compute_target=ws.DEFAULT_CPU_CLUSTER_NAME)\n",
" remote_diff.wait_for_completion()\n",
" \n",
" remote_diffs[dt] = remote_diff.get_result()\n",
" \n",
" for r in remote_diff.get_result():\n",
" if r.name not in r_measurements:\n",
" r_measurements.add(r.name)\n",
" if \"column_name\" in r.extended_properties and r.extended_properties[\"column_name\"] not in r_columns:\n",
" r_columns.add(r.extended_properties[\"column_name\"])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Parse And Present Remote Execution Results\n",
"\n",
"### Parse And List Column Level Diff Results\n",
"\n",
"<font size=2>Here will iteratively list all details per each measurement per column calculated.</font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from pandas import DataFrame\n",
"\n",
"r_dates = []\n",
"r_drift_metrics = {}\n",
"r_drift_contributions = {}\n",
"r_summary_contribute = {}\n",
"r_bottoms_contribute = []\n",
"\n",
"for dt, rst in remote_diffs.items():\n",
" r_dates.append(dt)\n",
" print(\"\\n---------------------------------------- Result of {} ----------------------------------------\".format(dt))\n",
" \n",
" daily_result, drift, daily_contribution = parse_result(rst, r_columns, r_measurements)\n",
" r_drift_metrics[dt] = drift\n",
" r_drift_contributions[dt] = daily_contribution\n",
"\n",
" sum_contribution = 0\n",
" r_bottoms_contribute.append(0)\n",
" for col, val in daily_contribution.items():\n",
" sum_contribution += val\n",
" r_summary_contribute[dt] = sum_contribution\n",
"\n",
" \n",
" display.display(pd.DataFrame(daily_result))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Present Dataset Level Diff (aka drift) In Graphs\n",
"\n",
"<font size=2><font size=2>The left graph presents dataset level difference for all compared baseline-target pairs, the right graph presents dataset level difference contribution for each column so that we know which column impacts more.</font></font>"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"show_diff(r_drift_metrics, r_dates, r_columns, r_drift_contributions, r_summary_contribute, r_bottoms_contribute)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Clean Resources Created"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"if del_cmpt == True:\n",
" try:\n",
" aml_compute.delete()\n",
" aml_compute.wait_for_completion()\n",
" except Exception as e:\n",
" if 'ComputeTargetNotFound' in e.message:\n",
" print(\"Compute target deleted.\")\n",
" del_cmpt = False"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Reference\n",
"\n",
"<font size=2>Detailed description of Dataset Diff attribute can be found at</font><br>\n",
"https://docs.microsoft.com/en-us/python/api/azureml-core/azureml.core.dataset(class)?view=azure-ml-py#diff-rhs-dataset--compute-target-none--columns-none-"
]
}
],
"metadata": {
"authors": [
{
"name": "davx"
}
],
"kernelspec": {
"display_name": "Python 3.6",
"language": "python",
"name": "python36"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.8"
},
"notice": "Copyright (c) Microsoft Corporation. All rights reserved. Licensed under the MIT License."
},
"nbformat": 4,
"nbformat_minor": 2
}

2
work-with-data/datasets/datasets-tutorial/datasets-diff/datasets-diff.ipynb
View File

@@ -24,7 +24,7 @@
"\n",
"<br>\n",
"<font size=2>\n",
" This notebook provides step by step instructions on how to compare two different datasets. It includes two parts：\n",
" This notebook provides step by step instructions on how to compare two different datasets. It includes two parts\u00ef\u00bc\u0161\n",
" <br>&nbsp;&nbsp;&nbsp;&nbsp;&#x2611; compare two datasets using local compute;\n",
" <br>&nbsp;&nbsp;&nbsp;&nbsp;&#x2611; compare two datasets remotely using Azure ML compute.\n",
"</font>"