hyiltiz / cones_mc Goto Github PK

main.m details typical usage: loading data, preparing calculations, running calculations and making some plots.

place the following two files containing your data in a folder such as peach/:

• stas.mat should contain a struct with fields: - stas(i).spikes the list of spike times of cell i - stas(i).spatial the spatial component of the STA of cell i

• cone_params.mat contains a struct with fields: - stimulus_variance : variance of each stim pixel color chanel (usually 1) - supersample : the integer number of cone positions per pixel width/height (usually 4) - colors : 3x3 matrix of cone color sensitivities - support_radius : radius of cone receptive field filter (usually 3.0) - repulsion_radii

see fullstas2stas.m for an example of how to make stas(i).spatial from raw data

the current main.m saves results from GREEDY.m, MCMC.m and CAST.m to the current folder.

a new folder is created by make_plots.m reflecting the name of the dataset; all appropriate plots are placed in it.

some plots are in .svg format; these files need evidence.png to live alongside them in the same folder; in order for these plots to be automatically converted to pdf, there must be a folder called batik/ containing a working version of Apache Batik in the same directory as main.m.

one of the plots for MCMC and CAST output is dancing_cones_movie.svg; this movie can be viewed by opening it in a modern browser.

most important files are commented; M2HTML is your friend