Skip to contents

This function loads spike raster data from a data frame or CSV file and converts each unique cell into a neuron object. The raster data frame must contain columns for cell identifier, spike time in milliseconds, and trial number. Optional metadata columns can also be included.

Usage

load.rasters.as.neurons(
  raster_df,
  bin_size = 10,
  min_duration = 0,
  max_displacement = 1e+09,
  sample_rt = 10000,
  time_cutoff = Inf
)

Arguments

raster_df

Data frame (or file name to csv importable as such), each row a spike; must have columns: cell, time_in_ms, trial; optional columns: recording_name, hemisphere, genotype, sex, region, age.

bin_size

Size of time bins in milliseconds (default: 10).

min_duration

Minimum trial duration in milliseconds (default: 0).

max_displacement

Max displacement of trial center in milliseconds (default: 1e9, i.e., "infinity").

sample_rt

Sample rate in the default unit for neuron objects, Hz (default: 1e4).

time_cutoff

Maximum time (in ms) to include spikes; spikes occurring after this time will be excluded (default: Inf).

Value

A list of neuron objects, one per unique cell in the raster data.