error in svm train

Error using svmtrain (line 267) TRAINING and Y must both contain at least one row with no missing values.

Error in signaltrain (line 213) SVMSTRUCT = svmtrain(Fsig, label1);

[Walter Roberson]

... pass in data that has at least one row that has no missing values?

[zara mansoor]

[zara mansoor]

there are NaNs in my data maybe thats why its howing this error how can I replace NaNs with zeros

[Walter Roberson]

Yes, you already posted that. Have you tried passing values in to svmtrain that have at least one row that does not have a missing value?

[zara mansoor]

I have converted NaNs to zeros bt now its giving error in svmclassify that-> Subscript indices must either be real positive integers or logicals.

[Walter Roberson]

You might run into problems if your label1 has any 0 values, unless label1 is type "logical". If you have some entries whose group is undefined then you should use label1 NaN for them -- but you will need to have at least one row whose group is not NaN.

[zara mansoor]

Thanks Walter I got it.