| Description |
PLOTSTATS(XDATA,YDATA) generate a nice figure for the statistical visualisation of datasets. YDATA is a cell array of vectors, each vector being a group of samples to test. XDATA is a cell array of the same size of YDATA, each element being a single number representing the abscissa of the corresponding sample group in YDATA. XDATA can be omitted. Example:
ydata = {
[ 1 2 3 4 5 6 7 8 9 ]
[ 2 3 4 5 ]
10*rand(10,1) };
xdata = {
1
2
3 };
ydata
ydata =
[ 1x9 double]
[ 1x4 double]
[10x1 double]
xdata
xdata =
[1]
[2]
[3]
plotstats(xdata,ydata)
With this default syntax, the command generate a 2-panel figure. The first one containing the raw data plot, the second one the box plot.
It is possible to extensively customize the generated plot. See inline help for details.
--------
EXAMPLES
--------
% Basic plot
x = {1 2 4};
y={ rand(50,1), rand(50,1), 0.7 + 0.5*rand(100,1)};
plotstats(x,y)
% With everything, outliers and fit (default is straight line)
x = {10 5 20};
y={ 2*rand(50,1), rand(50,1), [2.1 + 0.8*rand(100,1)' 0.2 0.25 0.1]};
labels = {'medium' 'low' 'high'};
fitopt.DoFit = true; % option structure for the fit panel
plotstats(x,y,'SeriesLabels',labels,'PlotHistogram',true,'PlotFit',true,'PlotFitOptions',fitopt)
% With no graphical output and a output argument, one can get only the result struct:
x = {10 5 20};
y={ 2*rand(50,1), rand(50,1), [2.1 + 0.8*rand(100,1)' 0.2 0.25 0.1]};
labels = {'medium' 'low' 'high'};
fitopt.DoFit = true; % option structure for the fit panel
S = plotstats(x,y,'SeriesLabels',labels,'PlotRawData',false,'PlotBox',false,'PlotFitOptions',fitopt) |
