matching word-frequency with word-ID in Imagesearch.py

[koreaccm]

Hello, @jesolem
I'm truly thankful for PCV sources.
Recently, I found that there might be a problem of indexing in Imagesearch.py
I just want to confirm whether the original code is right.

if we look into querying function, we can notice query() -------> candidate_from_histogram() -------> candidates_from_word().

def candidates_from_word(self,imword):
      im_ids = self.con.execute( "select distinct imid from imwords where wordid=%d" % imword).fetchall()

Meanwhile, I think the value indexed to imword table is not word-id, but word-frequency.

def add_to_index(self, imname, descr):
      ...
      imwords = self.voc.project(descr)
      nbr_words = imwords.shape[0]

      # link each word to image
      for i in range(nbr_words):
          word = imwords[ i ]
          # wordid is the word number itself
          self.con.execute("insert into imwords(imid,wordid,vocname) values (?,?,?)",  (imid,word,self.voc.name))

So, it seems that word-id and word-frequency are compared. Isn't it wrong?
I think the add_to_index() should be fixed as comparison between word-id and word-id like below.

def add_to_index(self, imname, descr):
      ...
      imwords1 = self.voc.project(descr)
      imwords2 = imwords1.nonzero()[0]
      nbr_words = imwords2.shape[0]

      # link each word to image
      for i in range(nbr_words):
          word = imwords2[ i ]
          # wordid is the word number itself
          self.con.execute("insert into imwords(imid,wordid,vocname) values (?,?,?)",  (imid,word,self.voc.name))
          # store word histogram for image
          # use pickle to encode NumPy arrays as strings
          self.con.execute("insert into imhistograms(imid,histogram,vocname) values (?,?,?)", (imid,pickle.dumps(imwords1),self.voc.name))