8000 GitHub - cabouman/gmcluster: EM Clustering Algorithm

More Web Proxy on the site http://driver.im/

Name		Name	Last commit message	Last commit date
Latest commit History 83 Commits
demo		demo
dev_scripts		dev_scripts
docs		docs
gmcluster		gmcluster
.gitignore		.gitignore
.readthedocs.yaml		.readthedocs.yaml
LICENSE		LICENSE
README.rst		README.rst
requirements.txt		requirements.txt
setup.py		setup.py

Repository files navigation

GMCluster

GMCluster: An Unsupervised Algorithm for Modeling Gaussian Mixtures

This is an EM-based clustering package for python that is based on the following C package: >https://engineering.purdue.edu/~bouman/software/cluster/

The documentation for this package is available here: >https://gmcluster.readthedocs.io/

Installing

Clone or download the repository and get inside:

git clone https://github.com/cabouman/gmcluster.git
cd gmcluster

Install the conda environment and package
1. Option 1: Clean install from dev_scripts
  *****You can skip all other steps if you do a clean install.****
  
  To do a clean install, use the command:
  cd dev_scripts source clean_install_all.sh cd ..
2. Option 2: Manual install
  1. Create conda environment:
    
    Create a new conda environment named gmcluster using the following commands:
    
    conda create --name gmcluster python=3.9 conda activate gmcluster
    
    Anytime you want to use this package, this gmcluster environment should be activated with the following:
    
    conda activate gmcluster
  2. Install the dependencies:
    
    To install the packages, use the following command.
    
    pip install -r requirements.txt
  3. Install gmcluster package:
    
    Use the following command to install the package.
    
    pip install .
    
    To allow editing of the package source while using the package, use
    
    pip install -e .
  4. Build the documentation:
    
    Use the following command to build the documentation.
    
    cd docs pip install -r requirements.txt make clean html cd ..

Running Demo(s)

You can validate the installation by running demo scripts.

cd demo
python demo_1.py

0