tm_sklearn_pipeline.py

import pandas as pd
from sklearn import metrics
from sklearn import preprocessing
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfTransformer
from sklearn.naive_bayes import MultinomialNB, GaussianNB
from sklearn.svm import LinearSVC
from sklearn.svm import SVC
from sklearn.linear_model import SGDClassifier
from sklearn.preprocessing import LabelEncoder
from sklearn.pipeline import Pipeline
from sklearn.feature_extraction import DictVectorizer
from sklearn.model_selection import GridSearchCV
import re
import nltk
import os
import glob
import numpy as np
from sklearn.datasets import load_files
from pprint import pprint
from time import time
import logging
logging.basicConfig(format='%(levelname)s:%(asctime)s:%(message)s', level=logging.INFO)


def experiment(text_clf_pipeline, train, test):

    print('pipline info:')
    for i, step in enumerate(text_clf_pipeline.get_params(deep=False)['steps']):
        print('step {}: {}'.format(i, step))

    text_clf_pipeline.fit(train.data, train.target)
    predicted = text_clf_pipeline.predict(test.data)
    print('Accuracy = {}'.format(np.mean(predicted == test.target)))
    print('classification report \n{}'.format(metrics.classification_report(y_pred=predicted,
                                                                            y_true=test.target)))
    print('confusion matrix \n{}'.format(metrics.confusion_matrix(y_pred=predicted,
                                                                            y_true=test.target)))
    f1 = 'f1-score: {0:.3f}'.format(metrics.f1_score(y_pred=predicted, y_true=test.target))
    print(f1)
    return f1


text_clf_pipeline1 = Pipeline([('vect', CountVectorizer(ngram_range=(4, 4),
                                                        min_df=25,
                                                        analyzer='char_wb')),
                              ('tfidf', TfidfTransformer()),
                              ('clf', MultinomialNB()), ])

text_clf_pipeline2 = Pipeline([('vect', CountVectorizer(ngram_range=(5, 5),
                                                        min_df=25,
                                                        analyzer='char_wb')),
                              ('tfidf', TfidfTransformer()),
                              ('clf', MultinomialNB()), ])

text_clf_pipeline3 = Pipeline([('vect', CountVectorizer(ngram_range=(6, 6),
                                                        min_df=25,
                                                        analyzer='char_wb')),
                              ('tfidf', TfidfTransformer()),
                              ('clf', MultinomialNB()), ])

text_clf_pipeline4 = Pipeline([('vect', CountVectorizer(ngram_range=(7, 7),
                                                        min_df=25,
                                                        analyzer='char_wb')),
                              ('tfidf', TfidfTransformer()),
                              ('clf', MultinomialNB()), ])

if __name__ == '__main__':
    wiki_train = load_files('ar_arz_wiki_corpus/train/', encoding='utf-8')
    wiki_test = load_files('ar_arz_wiki_corpus/test/', encoding='utf-8')
    text_clf_pipeline1 = Pipeline([('vect', CountVectorizer(ngram_range=(4, 4),
                                                            min_df=25,
                                                            analyzer='char_wb')),
                                   ('tfidf', TfidfTransformer()),
                                   ('clf', GaussianNB()), ])
    experiment(text_clf_pipeline1, wiki_train, wiki_test)
    result = {}

    # for i in range(4, 7):
    #     min_gram = i
    #     max_gram = i
    #     text_clf_pipeline = Pipeline([('vect', CountVectorizer(ngram_range=(min_gram, max_gram),
    #                                                         min_df=25,
    #                                                         analyzer='char_wb')),
    #                                ('tfidf', TfidfTransformer()),
    #                                ('clf', MultinomialNB()), ])
    #     f1 = experiment(text_clf_pipeline)
    #     result[(min_gram, max_gram)] = f1
    # for i in range(4, 7):
    #     min_gram = i
    #     max_gram = i+1
    #     text_clf_pipeline = Pipeline([('vect', CountVectorizer(ngram_range=(min_gram, max_gram),
    #                                                         min_df=25,
    #                                                         analyzer='char_wb')),
    #                                ('tfidf', TfidfTransformer()),
    #                                ('clf', MultinomialNB()), ])
    #     f1 = experiment(text_clf_pipeline)
    #     result[(min_gram, max_gram)] = f1
    # print(result)