# load library
import os 
import sys

# set working directory
os.chdir("D:\\my_python_workfile\\Thesis\\movie_review\\review_polarity\\txt_sentoken")

dataset_dir_name = os.getcwd()
dataset_dir_name

'D:\\my_python_workfile\\Thesis\\movie_review\\review_polarity\\txt_sentoken'

# load library
import numpy as np
from sklearn.datasets import load_files
from sklearn.cross_validation import train_test_split
from sklearn.feature_extraction.text import CountVectorizer,TfidfVectorizer
from sklearn.feature_extraction.text import TfidfTransformer
from sklearn.pipeline import Pipeline
from sklearn.grid_search import GridSearchCV
from sklearn.svm import LinearSVC
from sklearn.naive_bayes import MultinomialNB
from sklearn import metrics

# load data,and split into training/test set
movie_reviews = load_files(dataset_dir_name)
 # split 
doc_terms_train,doc_terms_test,doc_class_train,doc_class_test =  train_test_split(
        movie_reviews.data,movie_reviews.target,test_size = 0.2,random_state = None)
    

len(doc_class_train),len(doc_class_test),(movie_reviews.target_names)

#print("\n".join(movie_reviews.data[0].split("\n"))[:20])

(1600, 400, ['neg', 'pos'])

# build a vectorizer/classifier pipeline
pipeline = Pipeline([
        ('vect', TfidfVectorizer(min_df=3, max_df=0.95)),
        ('clf', LinearSVC(C=1000)),
    ])

# grid search
parameters = {
        'vect__ngram_range': [(1, 1), (1, 2)],
    }
grid_search = GridSearchCV(pipeline, parameters, n_jobs=-1)
grid_search.fit(doc_terms_train,doc_class_train)

print(grid_search.grid_scores_)

[mean: 0.83750, std: 0.01659, params: {'vect__ngram_range': (1, 1)}, mean: 0.85938, std: 0.01338, params: {'vect__ngram_range': (1, 2)}]

# y_predicted
y_predicted = grid_search.predict(doc_terms_test)

# report
print(metrics.classification_report(doc_class_test,y_predicted,
                                   target_names = movie_reviews.target_names))

             precision    recall  f1-score   support

        neg       0.85      0.85      0.85       188
        pos       0.87      0.86      0.87       212

avg / total       0.86      0.86      0.86       400

# confusion matrix
confusion_matrix = metrics.confusion_matrix(doc_class_test,y_predicted)
print(confusion_matrix)

[[160  28]
 [ 29 183]]