Remove unused variables and parameters for Soft Cosine Measure

docs/notebooks/soft_cosine_tutorial.ipynb

@@ -41,7 +41,6 @@
    "outputs": [],
    "source": [
     "from time import time\n",
-    "start_nb = time()\n",
     "\n",
     "# Initialize logging.\n",
     "import logging\n",
@@ -149,7 +148,7 @@
     "    raise ValueError(\"SKIP: You need to download the google news model\")\n",
     "    \n",
     "model = KeyedVectors.load_word2vec_format('/data/GoogleNews-vectors-negative300.bin.gz', binary=True)\n",
-    "similarity_matrix = model.similarity_matrix(corpus, dictionary)\n",
+    "similarity_matrix = model.similarity_matrix(dictionary)\n",
     "del model\n",
     "\n",
     "print('Cell took %.2f seconds to run.' % (time() - start))"
@@ -401,7 +400,7 @@
     "num_best = 10\n",
     "dictionary = corpora.Dictionary(scm_corpus)\n",
     "scm_corpus = [dictionary.doc2bow(document) for document in scm_corpus]\n",
-    "similarity_matrix = model.wv.similarity_matrix(scm_corpus, dictionary)\n",
+    "similarity_matrix = model.wv.similarity_matrix(dictionary)\n",
     "instance = SoftCosineSimilarity(scm_corpus, similarity_matrix, num_best=num_best)\n",
     "\n",
     "print('Cell took %.2f seconds to run.' %(time() - start))"

gensim/models/keyedvectors.py

@@ -560,19 +560,18 @@ def similar_by_vector(self, vector, topn=10, restrict_vocab=None):
         """
         return self.most_similar(positive=[vector], topn=topn, restrict_vocab=restrict_vocab)
 
-    def similarity_matrix(self, corpus, dictionary, threshold=0.0, exponent=2.0,
-                          nonzero_limit=100, dtype=REAL):
+    def similarity_matrix(self, dictionary, threshold=0.0, exponent=2.0, nonzero_limit=100,
+                          dtype=REAL):
         """Constructs a term similarity matrix for computing Soft Cosine Measure.
 
         Constructs a a sparse term similarity matrix in the `scipy.sparse.csc_matrix` format for
         computing Soft Cosine Measure between documents.
 
         Parameters
         ----------
-        corpus : list of lists of (int, float) two-tuples
-            A list of documents in the gensim document format.
         dictionary : gensim.corpora.Dictionary
-            A dictionary associated with the corpus.
+            A dictionary that specifies a mapping between words and the indices of rows and columns
+            of the resulting term similarity matrix.
         threshold : float, optional
             Only pairs of words whose embeddings are more similar than `threshold` are considered
             when building the sparse term similarity matrix. Defaults to `0.0`.

gensim/similarities/docsim.py

@@ -595,7 +595,7 @@ class SoftCosineSimilarity(interfaces.SimilarityABC):
     >>> # Construct a bag-of-words corpus, a dictionary, and a term similarity matrix.
     >>> dictionary = Dictionary(corpus)
     >>> corpus = [dictionary.doc2bow(document) for document in corpus]
-    >>> similarity_matrix = model.wv.similarity_matrix(corpus, dictionary)
+    >>> similarity_matrix = model.wv.similarity_matrix(dictionary)
     >>> index = SoftCosineSimilarity(corpus, similarity_matrix, num_best=10)
     >>> # Make a query.
     >>> query = 'Yummy! Great view of the Bellagio Fountain show.'