fight-churn
1. What's New : Support for Python 3.9-3.12! (April 2025)
Fighting Churn With Data just got added support for Python 3.10, 3.11 and 3.12! This is in addition to the original supported python, 3.9.
Python 3.13 is still not supported due to shap, which doesn't have precompiled wheels for Python 3.13 yet.
Note: If you try to updgrade packages, you will need to run pip with --force-reinstall or you may get errors about numpy not being compiled with the current version.
Please file issues if you notice problems with any supported python version, 3.9-3.12.
2. Not So New Anymore : ChurnSim White Paper & Advanced Simulation (August 2023)
New ChurnSim White Paper
There is now a white paper describing the inner workings of the Fighting Churn WIth Data churn simulation program:
- ChurnSim: A Customer Churn Behavioral Simulation System For Education and Analysis
- Note: It is not necessary to read the white paper to learn the churn fighting techniques from the book - the white paper is intended for advanced data scientists who want to create their own simulations.
New Advanced Simulation
Do you want to try out Fighting Churn With Data techniques with a more realistic data set? Try the new simulation of a Customer Relationship Management (CRM) product! It contains many realistic features including: multiple product plans, multi-user customer accounts, multiple billing periods, valued actions, and more! With these features you can try out ALL the code listings in Fighting Churn With Data and experience a great new challenge on your path to mastering churn.
To try the new simulation follow the setup instructions and add the argument --config-name crm5 to the churnsim.py command line, or set the argument schema='crm5' to the function
run_standard_simulation if you are running the simulation from a package
installation. The simulation will take an hour or so on a typical computer and produce around
30GB of data in your PostgreSQL database. See the Setup Instructions for more details.
3. What is this repository?
This is code for the book "Fighting Churn With Data: Science and strategy for keeping your customers"; the book serves as a detailed guide to the code. You can get more information at:
- https://www.manning.com/books/fighting-churn-with-data, the publisher (Manning Publications)'s page for the book
- http://www.fightchurnwithdata.com, the author's blog site
- https://www.twitch.tv/carl24k_datascience, the author's live stream
- The author's You Tube Channel, Fighting Churn With Data Science
This page contains the most up to date setup instructions.
4. Setup Instructions
WARNING TO PYPI USERS: None of the internal links in this document work on the pypi.org website! To use the internal links in this README view it on github.
Note from the Author
Before you can load data or run the code you have to do some setup on your system. If you never have done this before it may seem like a lot of work, and it kind of is, but this amount of setup is routine when you begin to work with a new technology. These setup instructions should work for most people using either shell Python, Jupyter Notebook, or an IDE, on either Mac or Windows. That said, I must apologize because I am neither an expert in PostgreSQL nor an expert in Python, but I am about to give a lot of advice on how to setup and use these technologies - if you find I am not doing things the best way, or just not how you would have done it, please be patient. If you want to make things better please help help out! :)
Pre-Requisite Setup
Python and PostgreSQL are required.
Code Setup
After the pre-requesisites, you have a choice on how to run the Fighting Churn With Data Code:
- Command Line Setup : The fastest and easiest way to get started, using a
pipPython package installation. This is great if you want to run the code and see the output, but you do not want to change the code or run it line by line. - Jupyter Notebook Setup : This is another way to get started quickly, if you already have Jupyter Notebooks installed.
- Developers IDE Setup : This method takes more time, but allows you to run the code line by line in a debugger, or change it. (If you are a professional developer planning to really use the code for your company, this is for you.)
5. Errata for the 1st Printing
This version of the README is for the the code released with the fightchurn pip package in September 2021. In this version there have been some modifications to the paths described in the first printing of Fighting Churn With Data to enable packaging the Python code. The following two differences summarize the changes:
- To run the code listings, the first printing of the book refers to the script
fight-churn/listings/py/run_churn_list.py: the path in the current code isfight-churn/fightchurn/run_churn_listing.py - To generate the data, the first printing of the book refers to scripts under
fight-churn/data-generation/: the path in the current code isfight-churn/fightchurn/churnsim
Complete details of the errata are described here.
Authors
- Carl Gold carl24k
License
This project is licensed under the MIT License - see the LICENSE.md file for details.
1 maintainer
Metadata
1 maintainer
Release files for fightchurn 1.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| fightchurn-1.2.0.tar.gz | 75.5 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| fightchurn-1.2.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 224.0 kB
Release files / fightchurn-1.2.0.tar.gz
| Download URL | fightchurn-1.2.0.tar.gz |
|---|---|
| Size | 75.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
8aa732ae13153f4b32748b0c3db910f23a2655701a94405534187be21cebf380
|
|
BLAKE2b-256 checksum How to use checksums |
be6584160303e85f5e2a8a803a1100c77d919f3774194d8699e53c692f34e749
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.9.18
|
Release files / fightchurn-1.2.0-py3-none-any.whl
| Download URL | fightchurn-1.2.0-py3-none-any.whl |
|---|---|
| Size | 148.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
a789731ab6be67b52adbb6d4717c373714f82f9d6c9ec1941ebe37aa530a159d
|
|
BLAKE2b-256 checksum How to use checksums |
40f7cb0f66d9fec9d8e530f11fe7c8d2709acb9f316e4efd2f6ecb436677b313
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.9.18
|
