Can anyone plz help regarding svmtrain in matlab?

I get these errors when I use svmtrain() in matlab.What to do?
Error using | Matrix dimensions must agree.
Error in svmtrain (line 265) nans = any(isnan(training)|isnan(groupIndex),2);

3 Comments

features=[1659,1486;1635,1534;1509,1228;113,676;712,1780;778,34;2409,946;502,1600;1137,1456;1845,934;22,520;1953,982;1617,910;1587,796;1767,676;2205,772;1287,802;2715,550;1863,586;1239,526;2055,778;2271,556;1863,886;2601,370;];
testset=[1269,988;802,1456;945,820;2619,814;2565,724;421,780;844,1768;430,1372;1863,1900;592,52;1617,1762;2391,82;2169,1576;850,472;1587,706;1437,976;1509,1556;1309,2284;1785,868;1869,856;2169,1186;1761,994;1725,862;1833,790;1833,538;];
labels=[ones(24,1),-ones(24,1)]
svmstruct=svmtrain(features,labels);
This is very petty and you can ignore it, but most of the community would appreciate the extra half second to type out 'please' haha. Good luck!

Answers (1)

A grouping variable must have the same number of observations (rows) as the table, data set array, or numeric array you are grouping. (Read here )
Assuming you have 24 data points in 2D space (i.e. with two features each) and 24 labels of which 12 are 1's and 12 are -1's.
labels = [ones(12,1);-ones(12,1)];
Hope this helps.

This question is closed.

Asked:

on 3 Aug 2014

Closed:

on 20 Aug 2021

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!