FieldTrip

Note that this reference documentation is identical to the help that is displayed in Matlab when you type “help ft_spikedensity”.

  FT_SPIKEDENSITY computes the spike density function of the spike trains by
  convolving the data with a window.
 
  Use as
    [sdf]          = ft_spike_density(cfg, data)
    [sdf, sdfdata] = ft_spike_density(cfg, data)
  
  If you specify one output argument, only the average and variance of
  spikedensityfunction across trials will be computed and individual trials are
  not kept. See cfg.winfunc below for more information on the specific use.
 
  Inputs:
    DATA should be organised in a structure as obtained from the
    FT_APPENDSPIKE or FT_SPIKESTATION_SPIKE2DATA or T_PREPROCESSING
    function.
 
  Configurations:
    cfg.timwin         = [begin end], time of the smoothing kernel (default = [-0.1 0.1])
                         If cfg.winfunc = @alphawin, cfg.timwin(1) will be set to 0.
    cfg.outputunit     = 'rate' (default) or 'spikecount'. This determines the physical unit
                         of our spikedensityfunction, either in firing rate or in
                         spikecount.
    cfg.winfunc        = (a) string or function handle, type of window to convolve with (def = @gauss).
                             Options should be set with cfg.winfuncopt
                         - @gauss (default)
                         - @alphawin, given by win = x*exp(-x/timeconstant)
                         - For standard window functions in the signal processing toolbox see
                           WINDOW.
                         (b) vector of length nSamples, used directly as window
    cfg.winfuncopt     = options that go with cfg.winfunc
                         For cfg.winfunc = @alpha: the timeconstant in seconds (default = 0.005s)
                         For cfg.winfunc = @gauss: the standard devision in seconds (default =
                                          1/4 of window duration in seconds)
                         For cfg.winfunc = @wname with @wname any standard window function
                                          see window opts in that function and add as cell array
    cfg.latency        = [begin end] in seconds, 'maxperiod' (default), 'minperiod',
                         'prestim'(t>=0), or 'poststim' (t>=0).
    cfg.vartriallen    = 'yes' (default) or 'no'.
                         'yes' - accept variable trial lengths and use all available trials
                          and the samples in every trial. Missing values will be ignored in
                          the computation of the average and the variance.
                         'no'  - only select those trials that fully cover the window as
                          specified by cfg.latency.
    cfg.trials         =  numeric or logical selection of trials (default = 'all')
    cfg.keeptrials     = 'yes' or 'no' (default). If 'yes', we store the trials in a matrix
                          in output SDF as well.
    cfg.fsample        = additional user input that can be used when input
                         is a SPIKE structure, in that case a continuous
                         representation is created using cfg.fsample
                         (default = 1000);
  Outputs:
    - SDF is a structure similar to TIMELOCK (output from FT_TIMELOCKANALYSIS) and can be used
      in FT_TIMELOCKSTATISTICS for example.
    - SDFDATA is a raw DATA type structure that can be used itself in all
    functions that support
      raw data input (such as FT_TIMELOCKANALYSIS, FT_FREQANALYSIS).