from sklearn.metrics import precision_recall_fscore_support
from sklearn.feature_extraction.text import TfidfVectorizer

param_grid = [
  {'C': [1, 10, 100, 1000], 'kernel': ['linear']},
  {'C': [1, 10, 100, 1000], 'gamma': [0.001, 0.0001], 'kernel': ['rbf']},
 ]
kf = cross_validation.KFold(len(totaldata), n_folds=5)
X = totaldata
y = totallabel
clf = GridSearchCV(SVC(C=1), param_grid, cv=5)

for train_index, test_index in kf:
    y_train, y_test = y[train_index], y[test_index]
    vectorizer = TfidfVectorizer(ngram_range=(1,2))
    #print train_index
    c_train = [con for ind, con in enumerate(content_list) if ind in train_index]
    X_train = vectorizer.fit_transform(c_train)
    c_test = [con for ind, con in enumerate(content_list) if ind in test_index]
    X_test = vectorizer.transform(c_test)
    clf.fit(X_train, y_train)
    y_pred = clf.predict(X_test)
    precision, recall, fbeta, support = precision_recall_fscore_support(y_test, y_pred, average='macro')
    print 'macro', precision, recall, fbeta, support
    precision, recall, fbeta, support = precision_recall_fscore_support(y_test, y_pred, average='micro')
    print 'micro', precision, recall, fbeta, support