src/python-sdk-v1/deploy_drift_detection.py

# Copyright (c) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License.

from obs.drift.drift_analysis_scheduler import execute
from datetime import datetime
from azure.ai.ml.constants import TimeZone
from dateutil import tz
from azure.ai.ml.entities import CronSchedule, ScheduleStatus
from azureml.core import Workspace
import argparse


def parse_args():
    parser = argparse.ArgumentParser(description="Drift Detection")
    parser.add_argument("-c", type=str, help="Compute Name", default = 'cpu-cluster')
    parser.add_argument("-b", type=str, help="Base table name", default='mlmonitoring')
    parser.add_argument("-t", type=str, help="target table name", default='mlmonitoring')    
    parser.add_argument("-bf", type=str, help="base starting date", default='12/15/2021')
    parser.add_argument("-bt", type=str, help="base ending date", default='03/01/2022')
    parser.add_argument("-tf", type=str, help="target starting date", default='04/01/2022')
    parser.add_argument("-tt", type=str, help="target ending date", default='5/15/2022')
    return parser.parse_args()

def main():

    args = parse_args()

    schedule_start_time = datetime.now(tz=tz.gettz("PACIFIC STANDARD TIME"))
    cron_schedule = CronSchedule(
        expression="*/10 * * * *", #provide a cron express to schedule the recurring execution of drift detection job
        start_time=schedule_start_time,
        time_zone=TimeZone.PACIFIC_STANDARD_TIME,
        status=ScheduleStatus.ENABLED,

    )

    target_dt_shift_step_size= "W"

    ws = Workspace.from_config()

    subscription_id = ws.subsription_id
    resource_group = ws.resource_group
    workspace = ws.name

    compute_name = args.c
    base_table_name = args.b
    target_table_name = args.t
    base_dt_from = args.bf
    base_dt_to = args.bt
    target_dt_from = args.tf
    target_dt_to= args.tt

    ml_client, job_name = execute(subscription_id=subscription_id,resource_group=resource_group,workspace=workspace, compute_name =compute_name, 
    base_table_name =base_table_name,target_table_name =target_table_name, base_dt_from =base_dt_from, base_dt_to= base_dt_to,target_dt_from=target_dt_from, 
    target_dt_to=target_dt_to, bin="7d", limit=100000, concurrent_run=False, drift_threshold =0.4,target_dt_shift_step_size=target_dt_shift_step_size,job_cron_schedule=cron_schedule)

if __name__ == "__main__":
    main()