classifier_pipeline.py

#This pipeline allow us to change the type of classifier:
def class_pipeline(features, class_ground, clf):
    #Call Shuffle splitter:
    ss = StratifiedShuffleSplit(n_splits=50)
    
    #sm = EditedNearestNeighbours()
    #[features, class_ground] = sm.fit_resample(features, class_ground)
    
    #Alocate variables:
    class_pred_all = []
    class_test_all = []
    f1_score_all = []

    #K fold cross validation shuffled:
    for train, test in ss.split(features,class_ground):
        #train/test split embedings:
        features_train = features[train,:]
        class_train = class_ground[train]
        features_test = features[test,:]
        class_test = class_ground[test]
        #Fit the model (clf is a typically used word to call any classifier):
        clf.fit(features_train,class_train)
        #Predict literature class from word embeddings:
        class_predicted = clf.predict(features_test)
        # Calculate F1 score and store all predicted and test classes:
        f1_score_all = np.append(f1_score_all,f1_score(class_test,class_predicted,average='weighted'))
        class_pred_all = np.append(class_pred_all,class_predicted)
        class_test_all = np.append(class_test_all,class_test)

    return class_pred_all, class_test_all, f1_score_all