Merge pull request #144 from commonsense/mini-changes

use a previously unused variable, fix documentation
commonsense · Dec 4, 2017 · 9930974 · 9930974
2 parents 21d7fab + de870b2
commit 9930974
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 2 deletions.
diff --git a/conceptnet5/vectors/cli.py b/conceptnet5/vectors/cli.py
@@ -249,7 +249,7 @@ def run_export(input_filename, output_filename, language):
 @click.option('-k', default=300, help="Number of columns to reduce to")
 def run_miniaturize(input_filename, extra_vocab_filename, output_filename, k):
     """
-    Save a smaller version of a frame, which includes frequent terms and doesn't include phrases.
+    Save a smaller version of a frame, which includes frequent terms and phrases.
     """
     frame = load_hdf(input_filename)
     other_frame = load_hdf(extra_vocab_filename)

diff --git a/conceptnet5/vectors/miniaturize.py b/conceptnet5/vectors/miniaturize.py
@@ -60,7 +60,8 @@ def miniaturize(frame, other_vocab=None, k=300, debias=True):
     smaller = frame.loc[vocab]
     U, _S, _Vt = np.linalg.svd(smaller, full_matrices=False)
     redecomposed = pd.DataFrame(U[:, :k], index=vocab, dtype='f')
-    redecomposed = de_bias_frame(redecomposed)
+    if debias:
+        redecomposed = de_bias_frame(redecomposed)
     mini = (redecomposed * 64).astype(np.int8)
     mini.sort_index(inplace=True)
     return mini