mirror of
https://github.com/Azure/MachineLearningNotebooks.git
synced 2025-12-20 09:37:04 -05:00
19 lines
426 B
Python
19 lines
426 B
Python
from pyspark.sql import SparkSession
|
|
|
|
import argparse
|
|
parser = argparse.ArgumentParser()
|
|
parser.add_argument("--input", default="")
|
|
parser.add_argument("--output", default="")
|
|
|
|
args, unparsed = parser.parse_known_args()
|
|
|
|
spark = SparkSession.builder.getOrCreate()
|
|
sc = spark.sparkContext
|
|
|
|
arr = sc._gateway.new_array(sc._jvm.java.lang.String, 2)
|
|
arr[0] = args.input
|
|
arr[1] = args.output
|
|
|
|
obj = sc._jvm.WordCount
|
|
obj.main(arr)
|