Installation:
Create and activate a new conda environment, with the twoppp package inside:
- Follow the instractions to download and install the twoppp package
Clone this repository:
git clone https://github.com/NeLy-EPFL/dn_networks
Downloading all data: Please download all files from the respsective Dataset on Harvard Dataverse as follows:
Make a base directory called DN_Networks. The code to decompress and analyse the data assumes a structure as follows:
data_summary_dir = "path/to/data/DN_Networks"- imaging data directory --> all data from the Optogenetics_Dfd_population_imaging dataset
data_summary_dir/Optogenetics_Dfd_population_imaging
- headless and predictions directory --> all data from the Optogenetics_headless_behavior dataset
data_summary_dir/Optogenetics_headless_behavior
- supplementary directory --> all data from the Supplementary_data dataset
data_summary_dir/Supplementary_data
Update folder names in params.py: Open the params.py file and update the directory as follows:
data_summary_dir = "path/to/data/DN_Networks"- Verify that the relative folder names for imaging_data_dir, headless_predictions_data_dir, other_data_dir are correct. (They should be in case you followed above instructions.)
Decompress the data Run the decompress_data.py script in this folder to assemble the data into its original folder structure.
Making all figures: Run the make_all_plots.py script in this folder to create all plots from experimental data.
Making all videos: Re-making all videos requires the raw data (~1TB) that is available upon request to the authors. The code can be inspected anyway in the make_all_videos.py script.