GENDATF Generation of fruit data set with 3 classes

    A = GENDATF(N,SF,SG)

 INPUT
       N    Number of generated samples or vector with
            number of samples per class (opt, def: N = [100 100 60])
       SF   Variance of the normal distribution for the banana shaped
            classes(opt, def: SF=1)
       SG   Variance of the gaussian shaped "stone" class (opt, def: SG=7)

 OUTPUT
       A    Generated data

 DESCRIPTION
 Generation of a 2-dimensional 3-class dataset A of N objects. The fruit
 class is composed of two classes, 'apple' and 'banana'. The classes are
 generated using the PRTools routine GENDATB. The third class 'stone'
 has a Gaussian shape with variance SG.

 GENDATF requires PRTools 4.1 (http://prtools.org)

 SEE ALSO
 GENDATB, GAUSS

gendatf is referenced in examples:

• kb5: ROC analysis on two-class problems: choosing an operating point