svmfit <- svm(y ~ ., data = train1, kernel = "linear", cost = 10, scale = FALSE)

WARNING: reaching max number of iterations