gpdscl

Genus Proximum Differentia Specifica Classification Language (GPDSCL)

gpdscl is a pure Python package for modelling faceted classifications with (here: freeplane-) mindmaps.

Versions

  • Version 0.26, 2021-02-16

Installation

Currently we suggest to work with gpdscl in a jupyter notebook within a standard anaconda or miniconda environment. (Students of jbusse may easily can install gpdscl into their personal data science lab.)

cd /home/myself/my/path/to/my/private/workspace
clone https://github.com/jbusse2/gpdscl

Files and directories included are e.g.:

  • (another nested) directory gpdscl: the according python package, including

    • src contains the most importand file mm2ttl.py

    • examples

  • mindmaps: place your own mindmaps here. Add this directory to your backup schedule.

  • backups: Each run of mm2ttl.py will place a time stamped backup file here by default. Delete these files regularly.

  • ontologies: output of mm2ttl.py

  • minimal_example.ipynb

Getting Started

Start jupyter notebook:

jupyter notebook start &

In jupyter move to your private workspace, then into the directory gpdscl. In jupyter open the file minimal_example.ipynb.