e1071 svm predicted probabilties are all 0.5

Question

My svm classifier model always predict 0.5 as probabilities.

svm.model <- svm(repeater ~ idRepeatBuyRatio + idTotalPurchase + c + d, data = trainData, cost = 100, gamma = 1)

svm.pred <- predict(svm.model, testData, probability = TRUE)
head(attr(svm.pred, "probabilities"))
    t   f
1 0.5 0.5
2 0.5 0.5
3 0.5 0.5
4 0.5 0.5
5 0.5 0.5
6 0.5 0.5

which is strange because the same call without probabilities actually makes different classifications:

> svm.pred <- predict(svm.model, testData, probability = FALSE)
> head(svm.pred)
1 2 3 4 5 6 
f t f t f f 
Levels: f t

Can someone explain what I am doing wrong?

Answer 1

The default value for probability in the svm function is FALSE. If you set this to TRUE then you ought to get the results you want.