v0.3

.idea/vcs.xml

@@ -0,0 +1,6 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="VcsDirectoryMappings">
+    <mapping directory="$PROJECT_DIR$" vcs="Git" />
+  </component>
+</project>

classify_news.py

@@ -0,0 +1,80 @@
+from sklearn.feature_extraction.text import TfidfVectorizer
+from stopWords import stopWrdList
+
+def getTrnVect():
+    # code to get the trained vectors
+
+    import yaml
+
+    str_trained_vect = open('trn_vect.vec', 'r').read().split('\n')
+
+    str_trained_vect.pop(len(str_trained_vect)-1)
+
+
+    trained_vect = []
+    for i in str_trained_vect:
+        trained_vect.append(yaml.load(i))
+
+
+    del str_trained_vect, i
+
+    return trained_vect
+
+
+def classify_news(document):
+    # code to vectorize news to classify
+
+    from similarityMeasures import cos_sim
+
+    vect_to_classify = []
+
+    news = open(document, 'r').read()
+
+    vect_to_classify.append(news)
+
+    stop_words = stopWrdList()
+
+    vectorizer = TfidfVectorizer(strip_accents='ascii', analyzer='word', stop_words=stop_words, max_features=100)
+
+    X = vectorizer.fit_transform(vect_to_classify)
+    vector = X.toarray()
+
+    trained_vectors = getTrnVect()
+
+    # get dim
+
+    len_vector = len(vector[0])
+    len_train = len(trained_vectors[0])
+
+    vector = list(vector[0])
+    if len_train > len_vector:
+        for i in range(len_train - len_vector):
+            vector.append(0)
+
+    sim_vect = []
+    for i in trained_vectors:
+        sim_vect.append(cos_sim(vector, i))
+
+
+    maxi = max(sim_vect)
+
+
+    x = 0
+    for i in sim_vect:
+        if i == maxi:
+            y = x
+        x = x + 1
+
+    part_neu_vect = 'This note has neutral emotions and it is related with the party'
+    part_neg_vect = 'This note has negative emotions and it is related with the party'
+    part_pos_vect = 'This note has positive emotions and it is related with the party'
+    cont_neu_vect = 'This note has neutral emotions and it is related with the opposition'
+    cont_neg_vect = 'This note has negative emotions and it is related with the opposition'
+    cont_pos_vect = 'This note has positive emotions and it is related with the opposition'
+    neut_neu_vect = 'This note has neutral emotions and it is not particularly related a political party'
+    neut_neg_vect = 'This note has negative emotions and it is not particularly related a political party'
+    neut_pos_vect = 'This note has positive emotions and it is not particularly related a political party'
+
+    results = [part_neu_vect, part_neg_vect, part_pos_vect, cont_neu_vect, cont_neg_vect, cont_pos_vect, neut_neu_vect, neut_neg_vect, neut_pos_vect]
+
+    print(results[y])

main.py

@@ -1,8 +1,8 @@
 
-from newsTrain import trainVect, flagger
+from newsTrain import saveTraining
+from classify_news import classify_news
 
+# saveTraining()
 
-sert = trainVect()
+classify_news('news_to_classify.txt')
-
+classify_news('news2.txt')
-for i in sert:
-    print(i)

newsTrain.py

@@ -2,7 +2,6 @@ from sklearn.feature_extraction.text import TfidfVectorizer
 from stopWords import stopWrdList
 from retEmoDict import emoDic
 from clust import clustering
-import operator
 
 def trainPre(word_array, dict):
 
@@ -178,7 +177,7 @@ def trainVect():
 
     stop_words = stopWrdList()
 
-    vectorizer = TfidfVectorizer(strip_accents='ascii', analyzer='word', stop_words=stop_words)
+    vectorizer = TfidfVectorizer(strip_accents='ascii', analyzer='word', stop_words=stop_words, max_features=100)
 
     X = vectorizer.fit_transform(corpus)
     vector = X.toarray()
@@ -237,6 +236,8 @@ def trainVect():
         neut_neg_vect = [vector[x] for x in neut_neg_ind]
         neut_pos_vect = [vector[x] for x in neut_pos_ind]
 
+############################################  1
+
     len1 = len(part_neu_vect)
     if len1 != 0:
         for a in range(len1):
@@ -251,18 +252,24 @@ def trainVect():
     else:
         part_neu_vect = []
 
+############################################  2
+
     len1 = len(part_neg_vect)
     if len1 != 0:
         for a in range(len1):
             tmp = part_neg_vect[0]
-            tmp = operate_on_Narray(part_neg_vect[0], tmp[a + 1], lambda x, y: x + y)
+            tmp = operate_on_Narray(part_neg_vect[0], tmp[a+1], lambda x, y: x + y)
+
+        tmp = operate_on_Narray(part_neg_vect[0], tmp[a+1], lambda x, y: x / len1)
 
-        tmp = operate_on_Narray(part_neg_vect[0], tmp[a + 1], lambda x, y: x / len1)
         part_neg_vect = list(tmp)
 
+
     else:
         part_neg_vect = []
 
+############################################  3
+
     len1 = len(part_pos_vect)
     if len1 != 0:
         for a in range(len1):
@@ -275,6 +282,8 @@ def trainVect():
     else:
         part_pos_vect = []
 
+############################################  4
+
     len1 = len(cont_neu_vect)
     if len1 != 0:
         for a in range(len1):
@@ -287,6 +296,8 @@ def trainVect():
     else:
         cont_neu_vect = []
 
+############################################  5
+
     len1 = len(cont_neg_vect)
     if len1 != 0:
         for a in range(len1):
@@ -299,6 +310,8 @@ def trainVect():
     else:
         cont_neg_vect = []
 
+############################################  6
+
     len1 = len(cont_pos_vect)
     if len1 != 0:
         for a in range(len1):
@@ -311,6 +324,22 @@ def trainVect():
     else:
         cont_pos_vect = []
 
+############################################  7
+
+    len1 = len(neut_neu_vect)
+    if len1 != 0:
+        for a in range(len1):
+            tmp = neut_neu_vect[0]
+            tmp = operate_on_Narray(neut_neu_vect[0], tmp[a + 1], lambda x, y: x + y)
+
+        tmp = operate_on_Narray(neut_neu_vect[0], tmp[a + 1], lambda x, y: x / len1)
+        neut_neu_vect = list(tmp)
+
+    else:
+        neut_neu_vect = []
+
+############################################  8
+
     len1 = len(neut_neg_vect)
     if len1 != 0:
         for a in range(len1):
@@ -324,6 +353,8 @@ def trainVect():
     else:
         neut_neg_vect = []
 
+############################################  9
+
     len1 = len(neut_pos_vect)
     if len1 != 0:
         for a in range(len1):
@@ -341,3 +372,12 @@ def trainVect():
 
     return [part_neu_vect, part_neg_vect, part_pos_vect, cont_neu_vect, cont_neg_vect, cont_pos_vect, neut_neu_vect, neut_neg_vect, neut_pos_vect]
 
+
+
+def saveTraining():
+
+    sert = trainVect()
+    trnVect = open('trn_vect.vec', 'w')
+
+    for i in sert:
+        trnVect.write(str(i) + '\n')

similarityMeasures.py

@@ -0,0 +1,45 @@
+"""
+Created on Mon Apr 17 09:34:40 2017
+functions to calculate the similarity measure of two real vectors
+@author: nlp
+"""
+# The cosine measure definition
+def cos_sim(vect1, vect2):
+    if (len(vect1) == len(vect2)):
+        vect3 = []
+        for x in range(0, len(vect1)):
+            vect3.append(0)
+            
+        for x in range(0, len(vect1)):
+            vect3[x] = vect1[x] * vect2[x]
+    
+        n1 = norm(vect1)
+        n2 = norm(vect2)
+    
+        return sum(vect3)/(n1*n2)
+        
+    else:
+       return 0
+       
+# Norm of vector
+def norm(vect):
+    import math as mth
+    vect1 = []
+    for x in range(0, len(vect)):
+        vect1.append(0)
+            
+    for x in range(0, len(vect)):
+        vect1[x] = vect[x] * vect[x]    
+            
+    return mth.sqrt(sum(vect1))
+
+# Jacard similarity
+def jac_sim(set_A,set_B):
+    if (str(type(set_A)) and str(type(set_B))) == "<class 'set'>":
+        if set_A == set_B:
+            return len(set_A & set_B)/len(set_A | set_B)
+        else:
+            return len(set_A & set_B)/len((set_A | set_B) - (set_A & set_B))
+    else:
+        print('One of the inputs not of type set')
+        

trn_vect.vec

@@ -0,0 +1,9 @@
+[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.029564870972714475, 0.0, 0.031681585307806945, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.085145070883776985, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.056251146398826481, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.073563944295505224, 0.0, 0.10753175525256822, 0.0, 0.0, 0.035183423033850185, 0.0, 0.0, 0.047392323741008761, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.085145070883776985, 0.0]
+[]
+[]
+[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.070657270308847969, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.026519396344811225, 0.0, 0.028418068131773094, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.083744798839379686, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.058615615346713827, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.07637428499714545, 0.0, 0.0, 0.0, 0.070657270308847969, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
+[]
+[0.10208139064065742, 0.0, 0.092124985572785187, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.13300880960778988, 0.0, 0.0, 0.0, 0.046184567743872848, 0.13300880960778988, 0.049491179049355211, 0.0, 0.0, 0.0, 0.0, 0.096826048928060432, 0.12305240453991766, 0.0, 0.0, 0.0, 0.0, 0.14584484823505761, 0.0, 0.14584484823505761, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.14584484823505761, 0.0, 0.0, 0.14584484823505761, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.087872356484313893, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.11491742926792517, 0.0, 0.0, 0.0, 0.0, 0.0839900103007927, 0.0, 0.13300880960778988, 0.05496155170329832, 0.0, 0.0, 0.0, 0.12305240453991766, 0.0, 0.0, 0.0, 0.13300880960778988, 0.0, 0.0, 0.0, 0.0]
+[0.0, 0.0, 0.044617190598828134, 0.0, 0.0, 0.0, 0.059595695487669048, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.022367717606043218, 0.0, 0.023969147510598116, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.064417697030794654, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.026618511934701741, 0.0, 0.0, 0.0, 0.0, 0.0, 0.064417697030794654, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
+[0.18807877008848775, 0.0, 0.0, 0.24506066350553238, 0.24506066350553238, 0.0, 0.0, 0.0, 0.0, 0.24506066350553238, 0.0, 0.0, 0.0, 0.085092264553030247, 0.0, 0.091184495307262525, 0.0, 0.0, 0.0, 0.0, 0.17839612176741071, 0.22671659111470846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.26871028605351621, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.26871028605351621, 0.0, 0.0, 0.0, 0.0, 0.26871028605351621, 0.0, 0.22671659111470846, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.26871028605351621, 0.0, 0.24506066350553238, 0.0, 0.0, 0.0, 0.24506066350553238, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.16189948656287695, 0.0, 0.0, 0.24506066350553238, 0.24506066350553238, 0.0, 0.0, 0.0, 0.0, 0.21172839263647156, 0.0, 0.0, 0.0, 0.0, 0.1547464992194269, 0.0, 0.0, 0.10126332509418293, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
+[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.066242832843992142, 0.0, 0.0, 0.0, 0.0, 0.010488531607570335, 0.030206347527019731, 0.011239464200115866, 0.033121416421996071, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0279452444280088, 0.0279452444280088, 0.0, 0.033121416421996071, 0.0, 0.0, 0.0, 0.033121416421996071, 0.033121416421996071, 0.0, 0.033121416421996071, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.033121416421996071, 0.0279452444280088, 0.033121416421996071, 0.033121416421996071, 0.0, 0.0, 0.0, 0.0, 0.033121416421996071, 0.033121416421996071, 0.0, 0.033121416421996071, 0.0, 0.0, 0.0, 0.033121416421996071, 0.033121416421996071, 0.033121416421996071, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.033121416421996071, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.030206347527019731, 0.0, 0.0, 0.0, 0.026097788677415228, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.012481787757288977, 0.0, 0.0, 0.0, 0.0, 0.0, 0.030206347527019731, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0279452444280088]