Predict Customer Churn

• Project Predict Customer Churn of ML DevOps Engineer Nanodegree Udacity

Project Description

Our main task in this prpject is to write clean and organized code that follows the best coding practices. What we're trying to do in this project is predict when customers might stop using a bank's services, like when they close their accounts or switch to another bank.

Here's what we're doing in simple terms:

Explore the Data: We start by looking at the information we have in our dataset, which is like a big collection of data. We want to understand it better before doing anything else.

Prepare the Data: To make our predictions work well, we need to get the data in good shape. We organize and change it a bit to create 19 different pieces of information that our prediction tools can use.

Train the Models: We're using two different methods to make predictions, like using two different types of tools. One is called "sklearn random forest," and the other is "logistic regression." We're testing both to see which one works better.

Figure out Important Information: We want to know which things are the most important for making predictions. We use a tool called the SHAP library to help us see which things have the biggest impact.

Save the Best Models: Once we figure out the best way to make predictions, we save those methods along with how well they work. This way, we can use them again later.

Also, we've made sure our code follows the rules for good coding. We formatted our code to meet the PEP8 standard, and it got a high score of over 9.4 for clean code from the pylint module. This shows that our code is well-organized and easy to read.

Files and data description

Folders

Data: Contains the dataset in CSV format. images eda: Contains the output of data exploration. results: Stores testing results. models: Contains saved models in .pkl format. logs: Holds logs generated during the testing of the churn_library.py file.

Project Files

churn_library.py: This is a Python script file. churn_notebook.ipnyb: This is a Jupyter Notebook file. requirements.txt: This file lists the required dependencies for the project.

Pytest Files test_churn_script_logging_and_tests.py: This file contains unit tests and configuration files for testing.

Running Files

To successfully run the project, ensure you have Python 3.8 installed, along with the necessary Python packages as specified in the requirements.txt file. To initiate the project, execute the script python churn_library.py from the project's root folder.

The functionality of the project script, churn_library.py, has been thoroughly tested using the pytest Python package.

Certainly, here's a shorter version of the testing and logging instructions in a single block:

To test and log the results of each function, follow these steps:

1. Open your terminal or command prompt.

2. Navigate to the project directory:

3. Launch IPython and run the script:

   ipython churn_script_logging_and_tests_solution.p

   This executes the script and logs the results to a .log file.

4. View the log file:

   After running the script, check the output.log file in the same directory for test results and logged information

Use the log file to verify function correctness and diagnose issues.