libtimeseries provides an integrative way to plot and manage behavioral and neural time course data. It is specialized in handling trial structure with multiple conditions and differen events.
see demo_timecourse_data_analysis.m for a demo with detailed comments.
- plot a PSTH (show raster + average across trials for each condition)
- save a PSTH
- load multiple PSTHs
- filter PSTHs
- combine PSTHs with different time range (e.g., homogenize x axis of PSTHs)
- plot individual or population PSTHs
- perform multi-dimensional population analyses
plot_timecourse returns a struct variable that contains essential information about the time course plot. The table below describes select fields of the psth struct.
name | size | description |
---|---|---|
x | [1 * # of timepoints] | time point of PSTHs |
mean | [# of groups * # of timepoints] | average activity for each group |
sem | [# of groups * # of timepoints] | standard error of mean for each group |
std | [# of groups * # of timepoints] | standard deviation for each group |
numel | [# of groups * # of timepoints] | # of valid trials for each group |
pDiff | [1 * # of timepoints] | p-values to test whether responses are same or different across groups |
pBaseDiff | [# of groups * # of timepoints] | p-values to test whether responses are same or different from baseline |
event | [1 * # of events] | a table containing the medians of events |
rate_rsp | [# of trials * # of timepoints] | trial-to-trial responses that was used to compute the averages for each group |
ginfo.grp_idx | [# of trials * 1] | group indice of individual trials |
ginfo.unq_grp_label | [# of groups * 1] | string labels for the group indice |