Some additional ngrams methods

movie.py

@@ -8,8 +8,8 @@
 
 def read_reviews():
     print "Reading and parsing files..."
-    pos_files = [ngrams.ngrams(1, open("pos/"+i).read()) for i in os.listdir("pos")]
-    neg_files = [ngrams.ngrams(1, open("neg/"+i).read()) for i in os.listdir("neg")]
+    pos_files = [ngrams.ngrams(3, open("pos/"+i).read()) for i in os.listdir("pos")]
+    neg_files = [ngrams.ngrams(3, open("neg/"+i).read()) for i in os.listdir("neg")]
     classes = [1] * len(pos_files) + [0] * len(neg_files)
     print "Creating matrix..."
     mat = ngrams.ngrams_to_matrix(pos_files + neg_files, classes)

ngrams.py

@@ -43,6 +43,12 @@ def ngrams(n, s):
     grams[" ".join(current)] += 1
     return grams
 
+def ngrams_range(b, e, s):
+    g = {}
+    for i in range(b, e+1):
+        g.update(ngrams(i, s))
+    return g
+
 def ngrams_to_matrix(grams, classes):
     print "a"
     keysets = [set(k) for k in grams]