Input and Output Names and Formats - BUNPC/Homer3 GitHub Wiki

Note: Many user functions take the output of other user functions as their input, and may return the inputs listed in the table below. hmrR_StimRejection, for instance takes and returns stim, excluding stims that fall in regions denoted by tIncMan or tIncAuto. Which table they are listed in below depends on _whether or not they are available before the processing stream is run, and where they appear in the resulting groupResults structure.

Inputs

Inputs are made available to the processing stream functions by Homer3 before the processing stream is run. These include data from SNIRF files such as signals, probe geometries, and stimulus onset times, as well as time points and channels manually excluded in Homer3's GUI.

Name	Format
data	data structure defined by the SNIRF file specification
stim	stim structure defined by the SNIRF file specification
probe	probe structure defined by the SIRF file specification
tIncMan	Cell array of 1D logical arrays. Each cell represents a data block, the array must consist of 1s and 0s and be equal in length to the time series data in the data block. 1s denote included time points, 0s denote excluded time points.
mlActMan	Cell array of 1D logical arrays. Each cell represents a data block, the array must consist of 1s and 0s and be equal in length to the number of channels in the data block. 1s denote included channels, 0s denote excluded channels.

Outputs

Outputs are generated by user functions and returned.

Name	Format
dod	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
dc	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
dodAvg	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
dcAvg	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
dodAvgStd	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
dcAvgStd	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
dodSum2	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
dcSum2	DataClass instance or data struct with the following fields: dataTimeSeries: an n-by-m array of n data points per m channel; time: an n-by-1 array of time stamps in seconds per data point; measurementList: a MeasListClass instance or measurementList struct with equivalent fields, see the SNIRF specification.
tHRF	time range for HRF model
nTrials	number of trials

Misc Outputs

Miscellaneous outputs are generated by user functions and returned in the ProcResultsClass.misc structure.

Name	Format
tIncAuto	n-by-1 logical array of 1s and 0s defining each of n time points as included or excluded from analysis, respectively.
tIncAutoCh	n-by-m logical array of 1s and 0s defining each of n time points as included or excluded from analysis per channel, m.
mlActAuto	An n-by-1 logical array of 1s and 0s defining each of n channels as active or inactive in analysis.
hmrstats	A struct with the following fields: beta_label {1 X # of regressors}: labels for each regressor modeled in GLM; tval {# of regressors X # of channels X # of Hb species}: t statistics that tests the hypothesis of whether any of the beta weights are statistically different than zero. e.g. if HRF is modeled by a set of consecutive gaussians, with a total number of nB gaussians, the first 1:nB entries (tval(1:nB,:,1) are t statistics for the beta values for HbO during the first condition. One can double check beta_label to verify; pval {# of regressors X # of channels X # of Hb species}: p-values that correspond to the t-statistics for the null hypothesis (tval); ml {# of channels X # of Hb species}: measurement list that provides the source and detector combination for each channel; tval_contrast {# of channels X # of Hb species}: t statistics that tests the hypothesis of whether the beta weights for any given condition(s) is statistically different than the beta weights for another condition(s); pval_contrast {# of channels X # of Hb species}: p-values that corresponds to the t-statistics for the contrast (tval_contrast); contrast {1 X # of conditions}: contrast vector. e.g. to contrast condition 2 to condition 3 in an experimental paradigm with four conditions, c_vector would be [0 1 -1 0].
dcNew	The model of the HRF with the residual. That is, it is the data y with the nuisance model parameters removed. {# of whole time series X # of Hb species X # of channels}
dcResid	The residual between the data y and the GLM fit {# of whole time series X # of Hb species X # of channels}
beta	Weights of the HRF regressors {# of coefficients X # of Hb species X # of channels X # of conditions}
R	The correlation coefficient of the GLM fit to the data {# of channels X # of Hb species}