Make load Method:

Parser & loader.

model_data_analysis.py

+import numpy as np
+import pprint
+import re
+import csv
+
+
+def load():
+    file_info = open ("info_model.txt", 'r')
+    if file_info == None:
+        print("Info File \'info_model.txt\' is absent")
+        return False
+
+    # parse info file
+    data = file_info.read().split("\n")
+    for i, dat in enumerate(data):
+        # words = [words.split(" ") for words in dat]
+        if i == 0:
+            rows = int(''.join(re.findall(r'\d',dat)))
+            # print(rows)
+        if i == 1:
+            columns = int(''.join(re.findall(r'\d',dat)))
+            # print(columns)
+        if i == 3:
+            file_name_matrix = re.match(r'^\w+\.csv', dat).group(0)
+            # print(file_name_matrix)
+        if i == 5:
+            file_name_dictionary = re.match(r'^\w+\.txt', dat).group(0)
+            # print(file_name_dictionary)
+        if i == 7:
+            file_name_tags = re.match(r'^\w+\.txt', dat).group(0)
+            # print(file_name_tags)
+
+    # read matrix
+    matrix = np.genfromtxt(file_name_matrix, delimiter=';')
+    # print(matrix)
+
+    # read tags !
+    tags = [item[:-1] for item in open(file_name_tags)]
+    # print(tags)
+
+    # read dictionary
+    dictionary = [word[:-1] for word in open(file_name_dictionary)]
+    # print(dictionary)
+    return rows, columns, dictionary, tags, matrix
+
+
+def save():
+    """ Just not used function. template from news_classify_tags.py """
+    import os
+
+    # save info data
+    file_info = open ("info_model.txt", 'w')
+    if file_info == None:
+        return False
+    file_info.write ("number of features = " + str (len (self.vectorizer.get_feature_names ())) + '\n')
+    file_info.write ("Quantity of tags = " + str (len (self.tags)) + '\n')
+    file_info.write ("\ndata_tag_frequency.csv  ---- Model data matrix --- " + \
+                     str (self.trained_model.shape) + "\n")
+    file_info.write ("\nDictionary.txt -- Total number words in  is---" + str (
+        len (self.vectorizer.get_feature_names ())) + '\n')
+    file_info.write ("\ntags.txt --- Total number is ---" + str (len (self.tags)) + '\n')
+    file_info.close ()
+
+    # save model matrix
+    file_tag_frequency = 'data_tag_frequency.csv'
+    if os.path.isfile (file_tag_frequency):
+        os.remove (file_tag_frequency)
+    file_data = open (file_tag_frequency, "w")
+    if file_data == None:
+        print ("Can't create data storage file")
+        return False
+    np.savetxt (file_tag_frequency, self.tag_frequency_matrix, delimiter=';', fmt='%1.4f')
+    print ("Data is saved into file: " + file_tag_frequency + " " + str (
+        int (os.stat (file_tag_frequency).st_size / 1024)) + 'kB')
+
+    # save dictionary
+    file_info_dictionaries = open ("Dictionary.txt", "w")
+    if file_info_dictionaries == None:
+        return False
+    for line in self.vectorizer.get_feature_names ():
+        file_info_dictionaries.write (line + "\n")
+    file_info_dictionaries.close ()
+
+    # save tags
+    file_info_tags = open ("tags.txt", "w")
+    if file_info_tags == None:
+        return False
+    for text1, text2 in self.tags:
+        file_info_tags.write (text1 + " " + text2.decode ('ascii', "ignore") + "\n")
+    file_info_tags.close ()
+    return True
+
+if __name__ == '__main__':
+    rows, columns, data_dictionary, tags, matrix = load()
+    if data_dictionary == None:
+        print("Something wrong whit data files. ")
+    # print(rows, columns, data_dictionary, tags, matrix )
+    np_matrix = np.asarray(matrix)
+    print(np_matrix)
+