changes TF-IDF search to cosine similarity from dot product

src/do_pca_on_tfidf.py

@@ -47,11 +47,11 @@ def query_worker(query, dtm_svd, dtm_svd_mat, vocab_norm, concentration = 10 ):
 
     # calculate the average TF-IDF score of the query over topics:
     #mean_query_score = np.sum(np.mean(query_weights, axis=0) * dtm_svd_mat, axis=1)
-    mean_query_score = cosine_similarity(np.reshape(query_weights, shape = (1, -1)), dtm_svd_mat)
+    mean_query_score = np.reshape(cosine_similarity(np.reshape(query_weights, shape = (1, -1)), dtm_svd_mat), shape=-1)
 
     sorted_df = pl.DataFrame(
         {
-            'score-tfidf': np.reshape(mean_query_score, shape=-1),
+            'score-tfidf': mean_query_score,
             'file':my_files
         }).sort("score-tfidf", descending = True).with_columns(pl.Series("rank-tfidf", [i + 1 for i in range(len(mean_query_score))]))