PopulationCoding
includes some functions for dimensionality reduction and other analysis of neurobehavioral data.
A few functions of particular relevance include:
- dimred.SVCA: SVCA, originally described by Stringer et al. 2019, with a few extra conveniences. Utilized for reliable dimensionality estimation in scaling_analysis
- predict.canonical_cov: canonical covariance analysis, utilized for prediction of neural activity from behvaior in scaling_analysis.predict.predict_from_behavior
- corr.sig_stim_corr: identify neurons significantly correlated to a stimulus according to the protocol we used in Demas et al. 2021
pip install PopulationCoding
Check out the full API in the documentation.
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.