comet-toolbox

Dynamic functional connectivity toolbox for multiverse analysis


License
MIT
Install
pip install comet-toolbox==0.4.4

Documentation

Comet - A toolbox for dynamic functional connectivity and multiverse analysis

DOI PyPI Codacy Badge Documentation Status Coverage Status

About the toolbox

Please refer to the documentation for detailed information about the toolbox. The following README will only provide a very brief overview.

Please also note that the package is in an early stage of development, with frequent changes being made. If you intend to use this package at this stage, please feel free to contact me via the email address in the pyproject.toml file. Some features are also not yet tested, so there will be bugs (the question is just how many).

Installation and usage

It is recommended to use a dedicated Python environment (e.g. through conda) to mitigate the risk of potential version conflicts. Installation is possible through the Python Package Index (PyPI) or from the source code in this repository:

pip install comet-toolbox

Usage of the toolbox is then possible through either the GUI:

comet-gui

or (for more versatile usage) through the scripting API. For this, demo scripts are provided as starting points:

Current features

Functional Connectivity Graph Analysis Multiverse Analysis
Continuous
  • Sliding Window Correlation
  • Jackknife Correlation
  • Flexible Least Squares
  • Spatial Distance
  • Temporal Derivatives
  • Phase Synchronization
  • Leading Eigenvector Dynamics
  • Wavelet Coherence
  • Edge-centric connectivity
State-Based
  • SW Clustering
  • Co-activation Patterns
  • Discrete HMM
  • Continuous HMM
  • Windowless
Static
  • Pearson Correlation
  • Partial Correlation
  • Mutual Information
Optimized implementation
  • Average Path Length
  • Global Efficiency
  • Nodal Efficiency
  • Small-World Sigma
  • Small-World Propensity
  • Matching Index
Standard Graph Functions
  • Threshold
  • Binarise
  • Symmetrise
  • Negative weights
  • ...
BCT Integration
  • All BCT functions can be
    used seamlessly fory
    multiverse analysis
  • Many BCT functions are available in the GUI
Simple Definition
  • Forking paths as
    python dictionary
  • Analysis pipeline template with decision points
Generation
  • Universes are created
    as individual scripts
  • Modular approach
Analysis
  • Individual universes
  • Entire multiverse (parallel)
Visualization
  • Multiverse summary
  • Multiverse as a network
  • Specification curve analysis

Current progress (roughly)

  • Large set of dFC methods and graph measures
  • Multiverse analysis framework
  • Graphical user interface
  • Documentation
  • Tutorials for an easy start
  • Complete testing suite
  • GUI bugfixes and optimization
  • Integrating over multiverse results
  • Dynamic community measures