MoSeq2-Viz: Visualization toolbox for MoSeq2

Latest version number: 0.5.0

Features

Below are the commands/functionality that moseq2-viz currently affords. They are accessible via CLI or Jupyter Notebook in moseq2-app.

Usage: moseq2-viz [OPTIONS] COMMAND [ARGS]...

Options:
  --version  Show the version and exit.  [default: False]
  --help     Show this message and exit.  [default: False]

Commands:
  add-group                       Change group name in index file given a...
  copy-h5-metadata-to-yaml        Copies metadata within an h5 file to a...
  get-best-model                  Returns the model with the closest median...
  make-crowd-movies               Writes movies of overlaid examples of the...
  plot-group-position-heatmaps    Plots position heatmaps for each group in...
  plot-scalar-summary             Plots a scalar summary of the index file...
  plot-stats                      Plots syllable usages with different...
  plot-transition-graph           Plots the transition graph depicting the...
  plot-verbose-position-heatmaps  Plots a position heatmap for each session...

Exclusive GUI Function

  get_groups_command(index_file)  Prints the given index file's current session metadata groupings.

Run any command with the --help flag to display all available options and their descriptions.

Documentation

MoSeq2 uses sphinx to generate the documentation in HTML and PDF forms. To install sphinx, follow the commands below:

pip install sphinx==3.0.3 sphinx_click==2.5.0
pip install sphinx-rtd-theme
pip install rst2pdf

All documentation regarding moseq2-extract can be found in the Documentation.pdf file in the root directory, an HTML ReadTheDocs page can be generated via running the make html in the docs/ directory.

To generate a PDF version of the documentation, simply run make pdf in the docs/ directory.

Prerequisites

To use this package, you must already have computed a pca_scores.h5 files, a trained model: model.p, an index file moseq2-index.yaml listing all your analyzed sessions and paths to their extracted .h5 files.

• The index file is generated when aggregating the results in moseq2-extract
• The pca_scores are generated via moseq2-pca.
• The model is generated via moseq2-model.

Example Outputs

Crowd Movie Example:

Usage Plot Example:

Alternative Sorting Examples:

Usages by Mutated Behaviors (most mutated to least)

Sorting Syllable Statistic in Descending Order (Speed shown)

Scalar Summary Example:

Position Heatmap Example:

Transition Graph Example:

Contributing

If you would like to contribute, fork the repository and issue a pull request.