Update get_hadiths.py

get_hadiths.py

@@ -47,24 +47,26 @@ class HadithSearch:
         except Exception as e:
             pass
 
-        embedding = self._get_embedding(user_input, model='text-embedding-ada-002')
-        self.data['similarities'] = self.data.embeding.apply(lambda x: self._cosine_similarity(x, embedding))
+        user_embedding = self._get_embedding(user_input, model='text-embedding-ada-002')
+        self.data['similarities'] = self.data.embeding.apply(lambda x: self._cosine_similarity(x, user_embedding))
 
-        results = self.data.sort_values('similarities', ascending=False).head(int(num_hadiths))
+        results = self.data.sort_values('similarities', ascending=False).head(int(num_hadiths)).copy()
         try:
             results.drop(columns=["id","hadith_id", "embeding"], inplace=True)
         except:
             pass
+        print(f"Number of hadiths to display: {num_hadiths}")
+        print(f"Shape of df: {str(results.shape)}")
         formatted_results = self._format_results(results.to_dict(orient="records"))
         return formatted_results
 
     def _format_results(self, results):
-        for r in results:
+        for result in results:
             formatted_output = ""
-            formatted_output += "### Source: " + str(r["source"]) + " | Chapter name : "+ str(r["chapter"]) +" | Chapter number: " + str(r["chapter_no"]) + " | Hadith number : " +  str(r["chapter_no"]) + "\n\n"
-            formatted_output += "Similarity with query: " + str(round(r["similarities"]*100,2)) + "%" +" | Chain index: " + str(r["chain_indx"]) + "\n\n"
-            formatted_output += "### Hadith content:" + "\n\n" + str(r["text_en"]) + "\n\n"
-            formatted_output += "Arabic version: \n\n" + str(r["text_ar"])
+            formatted_output += "### Source: " + str(result["source"]) + " | Chapter name : "+ str(result["chapter"]) +" | Chapter number: " + str(result["chapter_no"]) + " | Hadith number : " +  str(result["chapter_no"]) + "\n\n"
+            formatted_output += "Similarity with query: " + str(round(result["similarities"]*100,2)) + "%" +" | Chain index: " + str(result["chain_indx"]) + "\n\n"
+            formatted_output += "### Hadith content:" + "\n\n" + str(result["text_en"]) + "\n\n"
+            formatted_output += "Arabic version: \n\n" + str(result["text_ar"])
             formatted_output += "\n\n-----------------------------------------------------------------------------------------------------\n\n"            
         formatted_output = formatted_output.replace("`", "")
         return  formatted_output