Skip to content

Latest commit

 

History

History
41 lines (31 loc) · 1.31 KB

README.md

File metadata and controls

41 lines (31 loc) · 1.31 KB

Tridy

A pipeline for classifying binary dynamics on digraphs using closed neighbourhoods, proposed in:

An application of neighbourhoods in digraphs to the classification of binary dynamics

Pedro Conceição, Dejan Govc, Jānis Lazovskis, Ran Levi, Henri Riihimäki, and Jason P. Smith

Requirements

Python packages:

  • pyflagsercontain - see here
  • pyflagser - see here
  • pandas
  • numpy
  • subprocess
  • concurrent
  • os
  • sys
  • json
  • networkx
  • scipy
  • pickle
  • time

To run

Download the spike train data from here (or use your own), extract the file into the data folder and then run

    (cd data && python extract_data.py)

Edit any entries in the json file that need changing for your required parameters.

Run with

    python ./pipeline.py example.json

The results of the pipeline will be printed into "./results/classification_accuracies_parameter.json", where parameter is the featurisation parameter used.

Note that the example adjacency matrix is large and loaded as a full matrix in the code, using 8GB of memory, as such your system will need more than 8GB of memory to run the pipeline on this dataset.