SIGNALSTAT Computes and plots statistical characteristics of a signal, including the data histogram, a fitted normal distribution, a normal ditribution fitted on trimmed data, a boxplot, and the QQ-diagram. The estimates value are printed in a panel and can be read as output. Optionally, a topographic map (see TOPOPLOT) can be plotted. The boxplot and the Kolmogorov-Smirnov test require the MATLAB Statistics Toolbox.

Usage: >> signalstat( data )
>> signalstat( data, plotlab, dlabel, percent );
>> [M,SD,sk,k,med,zlow,zhi,tM,tSD,tndx,ksh] = signalstat( data, plotlab, dlabel, percent, dlabel2, map, chan_locs );

Inputs:
data   
data vector

Optional inputs:
plotlab   
1: default->plot | 0: ->no plot
dlabel   
A label for the data ([]: default->'Potential [�V]')
percent   
percentage of data to exclude for trimmed mean & SD ([]:default->5) Excluded is 'percent'/2 high % and 'percent'/2 low %
dlabel2   
A title label for the statistics table
map   
Data vector to be displayed as topographic map. If a single integer, only the corresponding electrode location is displayed
chan_locs   
name of an EEG electrode position file (See >> topoplot example for format). Can also be a structure (see >> help pop_editset)

Outputs:
M,SD   
mean and standard deviation
sk,k   
skewness and kurtosis
med   
median
zlow,zhi   
low and high 'percent/2'-Percentile ('percent/2'/100-Quantile)
tM,tSD   
trimmed mean and SD, removing datazhigh
tndx   
index of the data retained after trimming
ksh   
output flag of the Kolmogorov-Smirnov test at level p=0.05 0: data could be normally distributed; 1: data are not normally distributed -1: test could not be executed

Author: Luca Finelli, CNL / Salk Institute - SCCN, 2 August 2002

See also: pop_signalstat, qqdiagram, eeglab()

See the matlab file signalstat.m (may require other functions)

Back to functions