Added example script

jefflund · Mar 18, 2019 · f86cf89 · f86cf89
1 parent 28a323d
commit f86cf89
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 7 deletions.
diff --git a/ankura/topic.py b/ankura/topic.py
@@ -76,7 +76,7 @@ def _sparse_topic_word(corpus, K, V, z_attr):
 def pdists(corpus, theta_attr, metric='cosine'):
     D = len(corpus.documents)
     thetas = np.array([doc.metadata[theta_attr] for doc in corpus.documents])
-    dists = spatial.distance.pdist(thetas, metric)
+    dists = spatial.distance.pdist(thetas[:D], metric)
     for ij in np.argsort(dists, axis=None):
         i, j = ij // D, ij % D
         if i == j:

diff --git a/scripts/example_xref.py b/scripts/example_xref.py
@@ -8,11 +8,10 @@
 Q = ankura.anchor.build_cooccurrence(bible)
 anchors = ankura.anchor.doc_anchors(bible, Q, 3000)
 topics = ankura.anchor.recover_topics(Q, anchors)
-
 ankura.assign.variational(bible, topics, 'theta')
 
-for i, j in itertools.islice(ankura.topic.pdists(bible, 'theta'), 1000):
-    print(
-        bible.documents[i].metadata['verse'],
-        bible.documents[j].metadata['verse'],
-    )
+for i, j in itertools.islice(ankura.topic.pdists(bible, 'theta'), 5000):
+    verse_i = bible.documents[i].metadata['verse']
+    verse_j = bible.documents[j].metadata['verse']
+    correct = verse_j in bible.documents[i].metadata['xref-tske']
+    print(verse_i, verse_j, correct)