From 1bebe8676b3a07bf1bc2c9ffbfc879558cd5b990 Mon Sep 17 00:00:00 2001 From: Marko Kolarek Date: Tue, 11 Jun 2024 17:07:11 +0200 Subject: [PATCH] Add ExpAn to Projects --- content/projects/expan.md | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) create mode 100644 content/projects/expan.md diff --git a/content/projects/expan.md b/content/projects/expan.md new file mode 100644 index 0000000..c7de90a --- /dev/null +++ b/content/projects/expan.md @@ -0,0 +1,30 @@ ++++ +title = 'Python library for statistical analysis of A/B tests' ++++ + +## ExpAn: Enhancing A/B Testing with Automation and Standardization + +**Github link**: [https://github.com/zalando/expan](https://github.com/zalando/expan) + +**Overview**: A/B tests, or Randomized Controlled Trials, are crucial for optimizing business processes and user experience across various industries. To streamline the statistical analysis of these experiments, ExpAn (Experiment Analysis) was developed as a Python library. This library not only standardizes data structures but also paves the way for automation by being embedded into services. + +**Objective**: The primary goal was to formalize and encode the analysis process of A/B tests, facilitating further automation by integrating the library into broader services. + +**Role**: As one of the main contributors on this project, I was responsible for: +- Open-sourcing the project on Github. +- Packaging the ExpAn library using setup.py. +- Publishing the package to PyPI. +- Automating the deployment process with TravisCI. +- Performing robust tests on multiple Python versions and environments using tox, ensuring high test coverage with coverage. + +**Technology Stack**: +- **Language**: Python +- **Libraries**: + - **pandas**: For structuring and analyzing data + - **scipy**: For advanced statistical analysis + - **tox**: For testing + - **Sphinx**: For documentation + - **coverage**: For measuring test coverage + +The combination of these technologies and processes ensured a robust, automated workflow, enhancing the efficiency and reliability of A/B test analyses. +