This README provides instructions for replicating the results of the paper: **pyrichlet: A Python Package for Density Estimation and Clustering Using Gaussian Mixture Models**. Please read the following instructions before running the "replication_script.py" script on your machine:

* The replication of Sections 3 & 4 requires the following (base) functionalities to be available in your Python installation:

- pandas
- numpy
- matplotlib
- pyrichlet

* The replication of the package comparison and running times in Subsection 4.3 requires the following add-on scripts and installations:

- mixes
- subprocess to run the R-script "aux.R" within Python (additionally, a folder named "temp" is required).
- bayesmixpy

The (exact) library version numbers used in the original paper are provided in the file "requirements.txt". Though, all scripts are executable and give the same results also with higher version numbers. Note that bayesmixpy provided full support for Windows at the moment of the publication of the manuscript. Alternatively, you can also resort to the colab notebook "replication_script.ipynb".

The package bayesmixpy can be installed in Linux and MacOS by following the notebook "replication_script.ipynb". The binaries for git, cmake and gcc must be available from a  PATH directory. A Dockerfile is provided with an installation of python 2.11 and the bayesmixpy package.
