derluke / datarobotx-idp

Overview

Collection of unofficial idempotent DataRobot helpers. Ready for use in your orchestration tool of choice.

Installation

git clone https://github.com/datarobot-community/datarobotx-idp
cd datarobotx-idp
pip install -e .

Usage

import os
from datarobotx.idp.execution_environments import get_or_create_execution_environment

dr_endpoint = os.environ['DATAROBOT_ENDPOINT']
dr_token = os.environ['DATAROBOT_API_TOKEN']

env_id_1 = get_or_create_execution_environment(dr_endpoint, dr_token, "image #1")
env_id_2 = get_or_create_execution_environment(dr_endpoint, dr_token, "image #1")
assert env_id_1 == env_id_2

env_id_3 = get_or_create_execution_environment(dr_endpoint, dr_token, "image #2")
assert env_id_1 != env_id_3

Notes for contributors

Rules

1. Public functions must be idempotent
2. Function signatures must be type hinted (enforced by mypy)
3. Functions must have numpydoc-style docstrings (enforced by ruff)
4. Functions must either have a unit or integration test

Principles

1. Group code so it can be easily grokked and edited at the .py level
2. Minimize dependencies
3. Isolate dependencies by submodule to reduce risk of dependency conflicts