I have a collection of text documents. I\'ve been asked to show each document in tf-idf vector space and in ntc form and then, train a svm model based on documents\' vectors