SDFEATPLOT Interactive plot of per-class distributions for each feature SDFEATPLOT(DATA) SDFEATPLOT(DATA,options) INPUT DATA SDDATA set OPTIONS 'lab' Labels (property) used for drawing histograms (default: 'lab') 'absolute'Visualize absolute frequencies (default: 'relative') 'bins' Number of histogram bins (default: 30) If vector is specified, it is used as bin center definition (Data set with single feature is required) 'style' Specify line style of each class in a cell array 'maxsamples' Limit maximum number of samples from DATA used for estimating histograms (default: [] = use all samples) 'stem' Uses stem function instead of plot function (useful for histograms) DESCRIPTION SDFEATPLOT shows a plot of per-class distributions for each feature. Use up and down cursor keys to move between features. By default, current labels are used. SDFEATPLOT allows interactive selection of the threshold by clicking. Right-click opens a menu where decision above and below threshold may be specified and threshold may be saved to workspace as pipeline. Confusion matrix may be displayed in threshold mode. This helps to get a quick idea what classes would be separated. KEYBOARD SHORTCUTS up/down change feature 's' Switch between default plot drawing and stem drawing for histograms 'u' Use unique data values for x-axis bins and plot as stem 'a' Automatic grid (number of bins spread linearly over min/max range) 'x' Specify x-axis binning vector and redraw current feature (provide variable name in workspace) Example: enter vector such as -10:0.1:10 Threshold mode: 's' Save threshold as a pipeline 'c' Show confusion matrix 'n' Show normalized confusion matrix EXAMPLES Specifying line style for the three classes in data: sdfeatplot(data,'style',{'r-','r--','b-.'}) READ MORE http://perclass.com/doc/guide/visualization.html#sdfeatplot
sdfeatplot
is referenced in examples: