CODEWITHSUNDEEP
rmachine-learning
user3527482
user3527482

Reputation: 21

Trying to compute misclassifiction error for bagging classifier in R, getting error message

I'm trying to compute the misclassification error on my data for the bagging classifier. I've done it successfully for two other sets of data by doing it this way:

mydata.bagging <- bagging(V5~., data=mydata, mfinal=10)
mydata.bagging.pred <- predict.bagging(mydata.bagging,newdata=mydata)
mydata.bagging.pred$confusion
mydata.bagging.pred$error

And it's worked perfectly fine. However, when I try to do the same thing with my third data, I get an error message that says

#Error in cbind(yval2, yprob, nodeprob) : number of rows of matrices must match (see arg 2)

I'm assuming this is because the third data has fewer rows than the other 2, but how would I be able to fix this problem and find the misclassification error?

Upvotes: 2

Views: 5131

Answers (2)

jote
jote

Reputation: 21

I had similar problem and then figured it out.

It is possible that all of your left-hand side values (V5) are the same. The error is thrown as a saying that no decision can be made as it is too easy.

My source: http://kleinfelter.com/learning-r-painful-r-learnings

Upvotes: 2

tulipnl
tulipnl

Reputation: 51

After removing all 'NA', the problem has gone. Also, the first column has to be index column.

Upvotes: 0

Related Questions