Code in churn_library.py completes the process for solving the data science process including:

• EDA
• Feature Engineering (including encoding of categorical variables)
• Model Training
• Prediction
• Model Evaluation

Store result plots including at least one:

• EDA: Univariate, quantitative plot
• EDA: Univariate, categorical plot
• EDA: Bivariate plot
• Model: Store at least two models. Recommended using joblib and storing models with .pkl extension.

Use one-hot encoding or mean of the response to fill in categorical columns. Currently, the notebook does this in an inefficient way that can be refactored by looping. Make this code more efficient using the same method as in the notebook or using one-hot encoding. Tip: Creating a list of categorical column names can help with looping through these items and create an easier way to extend this logic

Each function in churn_script_logging_and_tests.py is complete with tests for the input function.

Also add requirements.txt with needed dependencies and libraries for simple install.

Each function in churn_script_logging_and_tests.py is complete with logging for if the function successfully passes the tests or errors. It also should:

• All log information should be stored in a .log file, so it can be viewed post the run of the script.
• The log messages should easily be understood and traceable that appear in the .log file.
• The README should inform a user how they would test and log the result of each function. (?)

Update functions to move constants to their own constants.py file, which can then be passed to the necessary functions, rather than being created as variables within functions.

To complete:

• The code is following the PEP 8 coding style
• The README file provides an overview of the project, the instructions to use the code
• All functions and files have document string