Code used to generate figures in "Revisiting efficient representations of space in hierarchical place field populations"
All code uses Python 3.8.10.
Code to simulate populations and compute Fisher information can be found in src/hierarchical_fisher.py
. This code also contains an object FisherTheory
, which provides analytical expressions for the discriminability under various conditions.
Use src/mp_simrunner.py
to run simulations. This runner natively spawns a separate process to run simulations for populations instantiated with different parameters to improve runtime. See src/hierarchical_fisher.py
for more details on how to run simulations. The notebook that computes the discriminability over the simulated populations uses files saved from mp_simrunner.py
. If interested in recreating any of the findings, swap the file names used in the notebooks with the ones you generate from running mp_simrunner.py
.
Place field analysis is run using data from Rich et al. 2014, data available here. All analysis code is in the respective notebook.