By: Mostafa Safaie^*, Joanna C. Chang^*, Junchol Park, Lee E. Miller, Joshua T. Dudman, Matthew G. Perich and Juan A. Gallego.

This repository includes code to reproduce the figures in Safaie & Chang et. al., Nature, 2022.

Create a conda environment with conda env create -f env.yml and activate the environment with conda activate cca.

Note that a version of MATLAB is needed for matlabengine, which is used for Fig S2. Comment out the matlabengine line if you do not have MATLAB installed. Otherwise, change the version of matlabengine according to the MATLAB version you have installed.

Each figure in the paper has an associated Jupyter notebook under /paper. Running the cells reproduces all of the panels.

For the RNN simulations associated with Figure 5 and Figure S10, first run the simulations:

cd rnn && bash rnn.sh

The code has been tested on Linux (Ubuntu >18.04).

MIT.

Code from the following sources have been copied and used here under /packages or /tools with slight modifications:

• config_manager from Sebastian Lee
• pyaldata from Neural Analysis
• dPCA from machenslab
• TME from Gamaleldin Elsayed

Associated publications are referenced in the paper.

Questions can be directed to the corresponding authors, as issues on this repository, or to Mostafa or Joanna.

This work was supported in part by:

• grant number H2020-MSCA-IF-2020-101025630 from the Commission of the European Union (M.S.)
• grant number 108908/Z/15/Z from the Wellcome Trust (J.C.C.)
• grant numbers NS053603 and NS074044 from the NIH National Institute of Neurological Disorders and Stroke (L.E.M.)
• grant chercheurs-boursiers en intelligence artificielle from the Fonds de recherche du Quebec Santé (M.G.P.)
• grant number EP/T020970/1 from the UKRI Engineering and Physical Sciences Research Council (J.A.G.)
• grant number ERC-2020-StG-949660 from the European Research Council (J.A.G.).