print("\nPredicting a 1v1 classifier for exciting :\n")
svm_exciting_1v1 = OneVsOneClassifier(LinearSVC(random_state=0)).fit(
X, exciting).predict(X)
print(svm_exciting_1v1)
print("\nPredicting a 1vrest classifier for exciting :\n")
svm_exciting_1vr = OneVsRestClassifier(LinearSVC(random_state=0)).fit(
X, exciting).predict(X)
print(svm_exciting_1vr)
print("\nGetting results for 1vrest classifier using decision_function :\n")
exciting_train = X[0:600]
exciting_test = X[601:949]
exciting_train_labels = exciting[0:600]
exciting_true = exciting[601:949]
exciting_pred = OneVsRestClassifier(LinearSVC(random_state=0)).fit(
exciting_train, exciting_train_labels).decision_function(exciting_test)
exciting_pred = exciting_pred.astype(int)
exciting_pred[exciting_pred < 0] = 0
exciting_pred[exciting_pred > 1] = 1
exciting_classes = ['class exciting', 'class not_exciting']
print("\nClassification report for exciting :\n")
print(
classification_report(exciting_true,
exciting_pred,
target_names=exciting_classes))
print("\nAccuracy report for exciting :\n")
print(accuracy_score(exciting_true, exciting_pred))
print("\nGetting probability matrix for all examples for exciting :\n")
svm_exciting_fit = SVC(kernel='linear', probability=True).fit(X, exciting)
svm_exciting_probs = svm_exciting_fit.predict_proba(X)
print(svm_exciting_probs)
