SOMptimised

An optimised Self Organising Map module based on sklearn-som with additional features.

Installation

SOMptimised is available for installation through pip by running the following command

pip install SOMptimised

Dependencies

Below are listed the dependencies required for the library to run with their latest version with which it was tested:

Library	Version	Installation (pip)	Installation (conda)
numpy	1.21.2	pip install numpy	conda install numpy
joblib	1.1.0	pip install joblib	conda install joblib
colorama	0.4.4	pip install colorama	conda install -c anaconda colorama

Tutorial/Example

A tutorial on how to use the SOM can be found below:

• Tutorial/Example
  • Preparing the data
  • Train the SOM to predict the iris class
  • Predict sepal width
  • Of the importance of normalisation
  • Set and get extra parameters
  • Save and load SOM