Updating imports and functions to work out the box

notebooks/Analyzing_Hacker_News_with_Six_Language_Understanding_Methods.ipynb

@@ -64,7 +64,7 @@
     "from annoy import AnnoyIndex\n",
     "import warnings\n",
     "from sklearn.cluster import KMeans\n",
-    "from bertopic._ctfidf import ClassTFIDF\n",
+    "from bertopic.vectorizers._ctfidf import ClassTfidfTransformer\n",
     "from sklearn.feature_extraction.text import CountVectorizer\n",
     "\n",
     "warnings.filterwarnings('ignore')\n",
@@ -889,8 +889,8 @@
     "# We'll need to embed the query using the same model that embedded the archive\n",
     "# so the query and archive are using the same embedding space.\n",
     "query_embed = co.embed(texts=[query],\n",
-    "                  model=\"small-20220425\", \n",
-    "                  truncate=\"RIGHT\").embeddings\n",
+    "                  model=\"small\", \n",
+    "                  truncate=\"LEFT\").embeddings\n",
     "\n",
     "# Retrieve the nearest neighbors\n",
     "similar_item_ids = search_index.get_nns_by_vector(query_embed[0],10,\n",
@@ -1086,8 +1086,8 @@
     "documents_per_topic = documents.groupby(['Topic'], as_index=False).agg({'Document': ' '.join})\n",
     "count_vectorizer = CountVectorizer(stop_words=\"english\").fit(documents_per_topic.Document)\n",
     "count = count_vectorizer.transform(documents_per_topic.Document)\n",
-    "words = count_vectorizer.get_feature_names()\n",
-    "ctfidf = ClassTFIDF().fit_transform(count).toarray()\n",
+    "words = count_vectorizer.get_feature_names_out()\n",
+    "ctfidf = ClassTfidfTransformer().fit_transform(count).toarray()\n",
     "words_per_class = {label: [words[index] for index in ctfidf[label].argsort()[-10:]] for label in documents_per_topic.Topic}\n",
     "df['cluster'] = classes\n",
     "df['keywords'] = df['cluster'].map(lambda topic_num: \", \".join(np.array(words_per_class[topic_num])[:]))"
@@ -1282,4 +1282,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}