scaling_analysis

scaling_analysis enables estimation of the reliable dimensionality of neuronal population dynamics and its scaling as a function of the number of sampled neurons, as described by Manley et al. Neuron 2024.

The most important functions include:

• svca.run_SVCA_partition samples a given number of neurons and performs shared variance component analysis (SVCA) using PopulationCoding.dimred.SVCA.

• predict.predict_from_behavior performs SVCA on a sampling of a specified number of neurons and then predicts the neural SVCs from behavioral variables.

Note that the analysis modules each contain a command line interface (CLI) which is described in the main() function within each module in the API.

Usage

pip install scaling_analysis

Check out the demos for examples of the analyses described in Manley et al. 2024.

Example datasets are freely available at https://doi.org/10.5281/zenodo.10403684.

Check out the full API in the documentation.

Citation

If you use this package, please cite the paper:

Manley, J., Lu, S., Barber, K., Demas, J., Kim, H., Meyer, D., Martínez Traub, F., & Vaziri, A. (2024). Simultaneous, cortex-wide dynamics of up to 1 million neurons reveal unbounded scaling of dimensionality with neuron number. Neuron. https://doi.org/10.1016/j.neuron.2024.02.011.

• Demo: 3x5mm Right Hemisphere FOV
  • SVCA on varying neuron numbers
    • Shuffling tests
    • The reliable dimensionality scales in an unbounded fashion with the number of randomly sampled neurons
  • Predictability of neural SVCs from behavior videography
    • Only the few largest neural SVCs are predictable from instantaneous behavior
      • The fraction of neural variance explained by behavior saturates at ~10,000 neurons
  • Autocorrelation timescales of neural SVCs
    • Neural SVCs exhibit a continuum of timescales
  • Spatial distribution of neural SVCs
    • Lower (behavior-related) neural SVCs are spatially clustered
• Demo: 5x6mm Bi-hemispheric FOV
  • SVCA on varying neuron numbers
    • Shuffling tests
    • The reliable dimensionality scales in an unbounded fashion with the number of randomly sampled neurons
  • Predictability of neural SVCs from behavior videography
    • Only the few largest neural SVCs are predictable from instantaneous behavior
      • The fraction of neural variance explained by behavior saturates at ~10,000 neurons
  • Autocorrelation timescales of neural SVCs
    • Neural SVCs exhibit a continuum of timescales
  • Spatial distribution of neural SVCs
    • Lower (behavior-related) neural SVCs are spatially clustered
• Demo: 1.2x1.2mm FOV
  • SVCA on varying neuron numbers
    • Shuffling tests
    • The reliable dimensionality scales in an unbounded fashion with the number of randomly sampled neurons
  • Predictability of neural SVCs from behavior videography
    • Only the few largest neural SVCs are predictable from instantaneous behavior
      • The fraction of neural variance explained by behavior saturates at ~10,000 neurons
  • Autocorrelation timescales of neural SVCs
    • Neural SVCs exhibit a continuum of timescales
  • Spatial distribution of neural SVCs
    • Lower (behavior-related) neural SVCs are spatially clustered

API:

• scaling_analysis package
  • Submodules
  • scaling_analysis.experiment module
    • Experiment
  • scaling_analysis.pca module
    • main()
    • run_pca_nneur()
  • scaling_analysis.plotting module
    • calc_var_expl()
    • log_log()
    • no_ticks()
    • plot_MIPs()
    • plot_neurons_behavior()
    • set_my_rcParams()
  • scaling_analysis.predict module
    • main()
    • predict_from_behavior()
  • scaling_analysis.predict_history module
    • main()
    • predict_from_behavior_history()
  • scaling_analysis.session_permutation module
    • main()
    • svca_across_sessions()
  • scaling_analysis.spatial module
    • gini_index()
    • hoyer_sparsity()
    • local_homogeneity()
    • reorder_svcs()
  • scaling_analysis.svca module
    • checkerboard_centers()
    • expanding_FOV()
    • main()
    • random_lateral_FOV()
    • run_SVCA_partition()
  • scaling_analysis.temporal module
    • autocorr()
    • compute_timescales()
    • find_svca_temporal()
    • fit_exponential_decay()
    • main()
  • scaling_analysis.utils module
    • correct_lag()
    • get_nneur()
  • scaling_analysis.visual module
    • compare_spont_vis_dims()
    • main()
  • Module contents

Indices and tables

• Index

• Module Index

• Search Page