Code implements the LN model used to describe spike trains of MEC neurons based on the animal's position, head direction, speed, and theta phase information. Used in Hardcastle et al., 2017.
Run_me.m is the main script. This will load the data from a single cell, fit the 15 LN models (as described in Hardcastle et al., 2017), select the best model according to a forward search procedure, and then plot the results. Details for each step can be found in run_me.m, and the scripts listed within run_me.m.
Updated versions of this code using splines can be found at https://github.com/GiocomoLab/spline-lnp-model. This version will implement an automated variable-fitting procedure, and will generally run faster with higher model fits. However, this version might be less user-friendly than the code in this repo.
Note - small updates and bug fixes have been made to the code. Questions can be directed to [email protected]