Analyzing multiple independent channelsAn ndata file may contain multiple data channels. For example, the file may have 2 or more channels, each containing spike trains recorded from a different neuron or a different electrode. Or, if the file was created from a model that has hundreds or many thousands of units in a population, there may be many channels with names like mt_0_0, mt_0_1, etc.
One could analyze each channel by using many 'nda' commands, as follows:
nda plaid120.nda model.nd out.pl chan mt_0_0 nda plaid120.nda model.nd out.pl chan mt_0_1 ...However, there are two simple, automated ways to do this:With the 'nda_pop' command, pop simply refers to the fact that there is a population of channels (more than one) to be analyzed:
- nda_pop - this command sequentially launches multiple independent nda commands, and is described below.
- chan_file - this parameter causes one nda command to loop over a list of channel names. This is described in the section on Common parameters.
nda_pop all_chan plaid120.nda model.nd out.plThis 'nda_pop' command will open the input .nd file and will find all of the channel names in the first trial. It will then issue separate "nda" commands (like those shown above) for each channel. A list of options for running nda_pop is as follows:Special nda_pop parameters may be added to an .nda file to improve the formatting of results for 'nda_pop', as follows:
- all_chan - use all channels
nda_pop all_chan [nda_file] [nd_file] [out_file]
- all_chan_prefix - use only channels whose name begins with a specified prefix
nda_pop all_chan_prefix [nda_file] [nd_file] [out_file] [prefix]
- all_chan_type - use only channels that have a certain channel type, e.g., 0 for spike times, 1 for floating point data.
nda_pop all_chan_type [nda_file] [nd_file] [out_file] [chan_type]
- all_chan_prefix_type - use only channels that have a certain channel type and whose name begins with the specified prefix.
nda_pop all_chan_prefix_type [nda_file] [nd_file] [out_file] [prefix] [chan_type]
- nda_pop_flag 1 - The default value is 0 if this parameter does not exist in the .nda script file. This can be used to inform the nda analyzer that this analysis is part of a population run. In this case, the nda analysis may (depending whether the analysis code detects and acts on this flag) decide to use the channel names to name any plots, or to include in any lines of statistical outputs that are appended to files.
- nda_pop_plotfile by_channel - The default value is "all" if this paramter is not present in the .nda script file. By setting this to by_channel, the analysis may (if it is implemnted) use the nda [out_file] as a prefix and create a distinct filename for each output plot by appending the channel name.
- Analysis types that currently look for these flags include:
- characterize_plaid
Future development.
- This type of population analysis could be interfaced with the look-up table analysis to customize the analysis for each channel.