- Project Predict Customer Churn of ML DevOps Engineer Nanodegree Udacity
This project aims to demonstrates the workflow of coding best practices for a standard data science task. This includes best practices on; general workflow, testing, logging, and PEP8 standards.
Running the below will install all dependencies necessary to run both python scripts.
pip install -r requirements.txt
Running the below will test each function from churn_library.py. During this process it will write log results in the ./logs directory.
ipython churn_script_logging_and_tests.py
Running the below will step through each function of the data science process. This includes; exploratory data analysis (EDA), feature transformations, model training and saving results.
ipython churn_library.py
This will generate:
- EDA plots in the ./images/eda/ directory
- training results in the ./images/results/ directory
- models that are saved in the ./models/ directory
- feature importances results in the ./images/results/ directory