istanbul-travel-guide.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Istanbul Travel Guide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "from dotenv import load_dotenv, dotenv_values\n",
    "\n",
    "# Load the .env file\n",
    "load_dotenv()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### create an index for the documents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "### create an index for the documents\n",
    "\n",
    "from langchain_openai import OpenAIEmbeddings\n",
    "from langchain_community.vectorstores import Chroma\n",
    "from langchain.text_splitter import RecursiveCharacterTextSplitter\n",
    "from langchain_community.document_loaders import PyPDFLoader\n",
    "\n",
    "file_paths = [\"C://Users//LENOVO//Desktop//LangGraph//datasets//istanbul-city-guide.pdf\",\n",
    "              \"C://Users//LENOVO//Desktop//LangGraph//datasets//istanbul-visitors-guide-2.pdf\",\n",
    "              \"C://Users//LENOVO//Desktop//LangGraph//datasets//istanbul-visitors-guide.pdf\"]\n",
    "\n",
    "embeddings = OpenAIEmbeddings(model=\"text-embedding-3-small\")\n",
    "\n",
    "docs = []\n",
    "\n",
    "for file_path in file_paths:\n",
    "    loader = PyPDFLoader(file_path)\n",
    "    docs.extend(loader.load())\n",
    "\n",
    "text_splitter = RecursiveCharacterTextSplitter(chunk_size=512,\n",
    "                                               chunk_overlap=64,\n",
    "                                               add_start_index=True)\n",
    "\n",
    "all_splits = text_splitter.split_documents(docs)\n",
    "\n",
    "vectorstore = Chroma.from_documents(documents=all_splits,\n",
    "                                    collection_name=\"guide-rag-chroma\",\n",
    "                                    embedding=embeddings\n",
    "                                   )\n",
    "\n",
    "retriever = vectorstore.as_retriever(search_type=\"similarity\", search_kwargs={\"k\": 3})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Document(metadata={'page': 14, 'source': 'C://Users//LENOVO//Desktop//LangGraph//datasets//istanbul-visitors-guide.pdf', 'start_index': 0}, page_content='13\\nHISTORY\\nTHe ONe aND ONl Y HagIa SOPHIa \\nThe Hagia Sophia, one of the \\nhistorical architectural wonders \\nof the world, has a very important place in the art world with its architecture, grandness, size and functionality. The Hagia Sophia, the biggest church constructed by the East Roman Empire in Istanbul, \\nhas been constructed three times \\nin the same location. The church \\nwas the place in which rulers \\nwere crowned, and it was also the biggest operational cathedral in \\nthe city throughout the Byzantine'),\n",
       " Document(metadata={'page': 37, 'source': 'C://Users//LENOVO//Desktop//LangGraph//datasets//istanbul-visitors-guide-2.pdf', 'start_index': 879}, page_content='acter.  Mosque elements such as a mimbar, a mihrab, preacher’s platform and wooden balustrades \\nwere added inside. Mahmut I had a library built here. The southern courtyard includes mauso-\\nleums of Murat III, Mehmet III, and Selim II. The original baptistery was converted into a mausoleum for Mustafa I and Sultan Ibrahim. Hagia Sophia opened as a museum in 1935.\\nVisiting hours: 9:00-16:30 - Closed on MondaysHAGIA SOPHIA\\nDominating the Sultanahmet Square and the city’s skyline, this'),\n",
       " Document(metadata={'page': 14, 'source': 'C://Users//LENOVO//Desktop//LangGraph//datasets//istanbul-visitors-guide.pdf', 'start_index': 1971}, page_content='The Hagia Sophia was converted \\ninto a museum by Mustafa Kemal Atatürk’s orders and has been functioning as one since 1935. The museum welcomes visitors from all around the world and is one of the most magnificent tourist attractions of Istanbul. The design and architecture of the museum is unbelievably gorgeous.')]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "retriever.invoke(\"Hagia Sophia\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### TranslateQuery"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Where is the Hagia Sophia mosque?\n",
      "What is the best traditional food to eat in Istanbul?\n"
     ]
    }
   ],
   "source": [
    "### TranslateQuery\n",
    "\n",
    "from langchain_core.output_parsers import StrOutputParser\n",
    "from langchain_openai import ChatOpenAI\n",
    "from langchain_core.prompts import ChatPromptTemplate\n",
    "\n",
    "# LLM\n",
    "llm = ChatOpenAI(model=\"gpt-4o-mini\", temperature=0)\n",
    "\n",
    "# Prompt\n",
    "system = \"\"\"Translate the question into English if the question is in another language. \\n\n",
    "If the question is in English, do nothing.\"\"\"\n",
    "translate_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\n",
    "            \"human\",\n",
    "            \"Here is the initial question: \\n\\n {question} \\n Translate the question into English.\",\n",
    "        ),\n",
    "    ]\n",
    ")\n",
    "\n",
    "turkish_question = \"Ayasofya cami nerededir?\"\n",
    "english_question2 = \"What is the best traditional food to eat in Istanbul?\"\n",
    "\n",
    "question_translater = translate_prompt | llm | StrOutputParser()\n",
    "print(question_translater.invoke({\"question\": turkish_question}))\n",
    "print(question_translater.invoke({\"question\": english_question2}))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### router"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "datasource='web_search'\n",
      "datasource='vectorstore'\n"
     ]
    }
   ],
   "source": [
    "### Router\n",
    "\n",
    "from typing import Literal\n",
    "\n",
    "from langchain_core.prompts import ChatPromptTemplate\n",
    "from langchain_core.pydantic_v1 import BaseModel, Field\n",
    "from langchain_openai import ChatOpenAI\n",
    "\n",
    "\n",
    "# Data model\n",
    "class RouteQuery(BaseModel):\n",
    "    \"\"\"\n",
    "    Route a user query to the most relevant datasource.\n",
    "    \"\"\"\n",
    "\n",
    "    datasource: Literal[\"vectorstore\", \"web_search\"] = Field(\n",
    "        ...,\n",
    "        description=\"Given a user question choose to route it to web search or a vectorstore.\",\n",
    "    )\n",
    "\n",
    "\n",
    "# LLM with function call\n",
    "llm = ChatOpenAI(model=\"gpt-4o-mini\", temperature=0)\n",
    "structured_llm_router = llm.with_structured_output(RouteQuery)\n",
    "\n",
    "# Prompt\n",
    "system = \"\"\"\n",
    "You are an expert at routing a user question to a vectorstore or web search.\n",
    "Vectorstore contains documents about the history of Istanbul, touristic and historical places of Istanbul, and food and travel tips for tourists.\n",
    "Use vectorstore for questions on these topics. If the question is about transportation, weather, and other things, use web search.\n",
    "\"\"\"\n",
    "\n",
    "route_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\"human\", \"{question}\"),\n",
    "    ]\n",
    ")\n",
    "\n",
    "question_router = route_prompt | structured_llm_router\n",
    "\n",
    "print(\n",
    "    question_router.invoke(\n",
    "        {\"question\": \"What is the weather like in Istanbul?\"}\n",
    "    )\n",
    ")\n",
    "\n",
    "print(\n",
    "    question_router.invoke(\n",
    "        {\"question\": \"Where is the Hagia Sophia located?\"}\n",
    "        )\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Retrieval Grader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "binary_score='yes'\n",
      "binary_score='no'\n"
     ]
    }
   ],
   "source": [
    "### Retrieval Grader\n",
    "\n",
    "# Data model\n",
    "class GradeDocuments(BaseModel):\n",
    "    \"\"\"Binary score for relevance check on retrieved documents.\"\"\"\n",
    "\n",
    "    binary_score: str = Field(\n",
    "        ...,\n",
    "        description=\"Documents are relevant to the question, 'yes' or 'no'\"\n",
    "    )\n",
    "\n",
    "\n",
    "# LLM with function call\n",
    "llm = ChatOpenAI(model=\"gpt-4o-mini\", temperature=0)\n",
    "structured_llm_grader = llm.with_structured_output(GradeDocuments)\n",
    "\n",
    "# Prompt\n",
    "system = \"\"\"You are a grader assessing relevance of a retrieved document to a user question. \\n \n",
    "If the document contains keyword(s) or semantic meaning related to the user question, grade it as relevant. \\n\n",
    "It does not need to be a stringent test. The goal is to filter out erroneous retrievals. \\n\n",
    "Give a binary score 'yes' or 'no' score to indicate whether the document is relevant to the question.\"\"\"\n",
    "grade_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\"human\", \"Retrieved document: \\n\\n {document} \\n\\n User question: {question}\"),\n",
    "    ]\n",
    ")\n",
    "\n",
    "retrieval_grader = grade_prompt | structured_llm_grader\n",
    "\n",
    "question = \"Where is the Hagia Sophia located?\"\n",
    "\n",
    "docs_relevant = retriever.invoke(question)\n",
    "doc_txt = docs_relevant[1].page_content\n",
    "print(retrieval_grader.invoke({\"question\": question, \"document\": doc_txt}))\n",
    "\n",
    "question_2 = \"What are LeBron James' stats for 2023-2024 NBA Regular season?\"\n",
    "\n",
    "docs_relevant_2 = retriever.invoke(question)\n",
    "doc_txt_2 = docs_relevant[1].page_content\n",
    "print(retrieval_grader.invoke({\"question\": question_2, \"document\": doc_txt_2}))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Generate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The Hagia Sophia is located in Sultanahmet Square, which is in the heart of Istanbul, Turkey. This iconic structure is one of the most significant historical and architectural wonders of the world. Originally built as a cathedral during the Byzantine Empire, it has served various purposes throughout its history, including being a mosque and, more recently, a museum. The Hagia Sophia is renowned for its stunning architecture and grand design, making it a must-visit attraction for tourists from around the globe.\n"
     ]
    }
   ],
   "source": [
    "### Generate\n",
    "\n",
    "from langchain import hub\n",
    "from langchain_core.output_parsers import StrOutputParser\n",
    "\n",
    "# Prompt\n",
    "prompt = hub.pull(\"e-ticaret-qa-prompt/istanbul-guide-rag\")\n",
    "\n",
    "# LLM\n",
    "llm = ChatOpenAI(model_name=\"gpt-4o-mini\", temperature=0)\n",
    "\n",
    "\n",
    "# Post-processing\n",
    "def format_docs(docs):\n",
    "    return \"\\n\\n\".join(doc.page_content for doc in docs)\n",
    "\n",
    "\n",
    "# Chain\n",
    "rag_chain = prompt | llm | StrOutputParser()\n",
    "\n",
    "# Run\n",
    "generation = rag_chain.invoke({\"context\": docs_relevant, \"question\": question})\n",
    "print(generation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hallucination Grader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradeHallucinations(binary_score='yes')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Hallucination Grader\n",
    "\n",
    "# Data model\n",
    "class GradeHallucinations(BaseModel):\n",
    "    \"\"\"Binary score for hallucination present in generation answer.\"\"\"\n",
    "\n",
    "    binary_score: str = Field(\n",
    "        description=\"Answer is grounded in the facts, 'yes' or 'no'\"\n",
    "    )\n",
    "\n",
    "\n",
    "# LLM with function call\n",
    "llm = ChatOpenAI(model=\"gpt-4o-mini\", temperature=0)\n",
    "structured_llm_grader = llm.with_structured_output(GradeHallucinations)\n",
    "\n",
    "# Prompt\n",
    "system = \"\"\"You are a grader assessing whether an LLM generation is grounded in / supported by a set of retrieved facts. \\n \n",
    "     Give a binary score 'yes' or 'no'. 'Yes' means that the answer is grounded in / supported by the set of facts.\"\"\"\n",
    "hallucination_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\"human\", \"Set of facts: \\n\\n {documents} \\n\\n LLM generation: {generation}\"),\n",
    "    ]\n",
    ")\n",
    "\n",
    "hallucination_grader = hallucination_prompt | structured_llm_grader\n",
    "hallucination_grader.invoke({\"documents\": docs, \"generation\": generation})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Answer Grader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GradeAnswer(binary_score='yes')"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Answer Grader\n",
    "\n",
    "# Data model\n",
    "class GradeAnswer(BaseModel):\n",
    "    \"\"\"Binary score to assess answer addresses question.\"\"\"\n",
    "\n",
    "    binary_score: str = Field(\n",
    "        description=\"Answer addresses the question, 'yes' or 'no'\"\n",
    "    )\n",
    "\n",
    "\n",
    "# LLM with function call\n",
    "llm = ChatOpenAI(model=\"gpt-4o-mini\", temperature=0)\n",
    "structured_llm_grader = llm.with_structured_output(GradeAnswer)\n",
    "\n",
    "# Prompt\n",
    "system = \"\"\"You are a grader assessing whether an answer addresses / resolves a question \\n \n",
    "     Give a binary score 'yes' or 'no'. Yes' means that the answer resolves the question.\"\"\"\n",
    "answer_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\"human\", \"User question: \\n\\n {question} \\n\\n LLM generation: {generation}\"),\n",
    "    ]\n",
    ")\n",
    "\n",
    "answer_grader = answer_prompt | structured_llm_grader\n",
    "answer_grader.invoke({\"question\": question, \"generation\": generation})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Question Re-writer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'What is the location of the Hagia Sophia?'"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### Question Re-writer\n",
    "\n",
    "# LLM\n",
    "llm = ChatOpenAI(model=\"gpt-3.5-turbo-0125\", temperature=0)\n",
    "\n",
    "# Prompt\n",
    "system = \"\"\"You a question re-writer that converts an input question to a better version that is optimized \\n \n",
    "     for vectorstore retrieval. Look at the input and try to reason about the underlying semantic intent / meaning.\"\"\"\n",
    "re_write_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\n",
    "            \"human\",\n",
    "            \"Here is the initial question: \\n\\n {question} \\n Formulate an improved question.\",\n",
    "        ),\n",
    "    ]\n",
    ")\n",
    "\n",
    "question_rewriter = re_write_prompt | llm | StrOutputParser()\n",
    "question_rewriter.invoke({\"question\": question})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Output guardrails"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "binary_score='yes'\n",
      "binary_score='no'\n"
     ]
    }
   ],
   "source": [
    "### Output guardrails\n",
    "\n",
    "from langchain_core.output_parsers import StrOutputParser\n",
    "from langchain_openai import ChatOpenAI\n",
    "from langchain_core.prompts import ChatPromptTemplate\n",
    "from langchain_core.pydantic_v1 import BaseModel, Field\n",
    "\n",
    "# Data model\n",
    "class EvaluateAnswer(BaseModel):\n",
    "    \"\"\"Binary score to assess answer addresses question.\"\"\"\n",
    "\n",
    "    binary_score: str = Field(\n",
    "        description=\"Answer addresses the question, 'yes' or 'no'\"\n",
    "    )\n",
    "\n",
    "# LLM with function call\n",
    "llm = ChatOpenAI(model=\"gpt-4o-mini\", temperature=0)\n",
    "structured_llm_response_evaluator = llm.with_structured_output(EvaluateAnswer)\n",
    "\n",
    "# Prompt\n",
    "system = \"\"\"You are an expert in checking that the answers produced do not contain statements that violate human rights.\n",
    "Check whether the outputs contain expressions that violate human rights, glorify violence, racism, sexism, etc. Give a binary score \"yes\" or \"no\".\n",
    "Yes means that the answer contains these type of statements.\"\"\"\n",
    "evaluate_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\n",
    "            \"human\",\n",
    "            \"Here is the generated response. \\n\\n {generation} \\n Check the generated response.\",\n",
    "        ),\n",
    "    ]\n",
    ")\n",
    "\n",
    "statement_1 = \"Black, gay men are a danger to society.\"\n",
    "statement_2 = \"Istanbul is the most crowded city in Turkey.\"\n",
    "\n",
    "answer_evaluator = evaluate_prompt | structured_llm_response_evaluator\n",
    "print(answer_evaluator.invoke({\"generation\": statement_1}))\n",
    "print(answer_evaluator.invoke({\"generation\": statement_2}))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### tavily web search tool"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "### tavily web search\n",
    "\n",
    "from langchain_community.tools.tavily_search import TavilySearchResults\n",
    "\n",
    "web_search_tool = TavilySearchResults(k=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'url': 'https://www.weatherapi.com/',\n",
       "  'content': \"{'location': {'name': 'Istanbul', 'region': 'Istanbul', 'country': 'Turkey', 'lat': 41.02, 'lon': 28.96, 'tz_id': 'Europe/Istanbul', 'localtime_epoch': 1724665871, 'localtime': '2024-08-26 12:51'}, 'current': {'last_updated_epoch': 1724665500, 'last_updated': '2024-08-26 12:45', 'temp_c': 31.1, 'temp_f': 88.0, 'is_day': 1, 'condition': {'text': 'Partly cloudy', 'icon': '//cdn.weatherapi.com/weather/64x64/day/116.png', 'code': 1003}, 'wind_mph': 16.1, 'wind_kph': 25.9, 'wind_degree': 40, 'wind_dir': 'NE', 'pressure_mb': 1010.0, 'pressure_in': 29.83, 'precip_mm': 0.0, 'precip_in': 0.0, 'humidity': 46, 'cloud': 25, 'feelslike_c': 32.3, 'feelslike_f': 90.1, 'windchill_c': 29.8, 'windchill_f': 85.6, 'heatindex_c': 30.4, 'heatindex_f': 86.6, 'dewpoint_c': 16.3, 'dewpoint_f': 61.4, 'vis_km': 10.0, 'vis_miles': 6.0, 'uv': 7.0, 'gust_mph': 16.7, 'gust_kph': 26.9}}\"},\n",
       " {'url': 'https://www.meteoprog.com/weather/Istanbul/month/august/',\n",
       "  'content': 'İstanbul (Turkey) weather in August 2024 ☀️ Accurate weather forecast for İstanbul in August ⛅ Detailed forecast By month Current temperature \"near me\" Weather news ⊳ Widget of weather ⊳ Water temperature | METEOPROG ... 26 August +28 °+23° 27 August +28 ...'},\n",
       " {'url': 'https://www.easeweather.com/europe/turkey/istanbul/august',\n",
       "  'content': 'Until now, August 2024 in Istanbul closely aligns with the historical average temperatures, with a minimal deviation of 0 ° C.. The forecast for the next days in Istanbul predicts the temperature to closely align with the historical average at 29.3 ° C.. In general, the average temperature in Istanbul at the beginning of August is 30.3 ° C.As the month progressed, temperatures tended to ...'},\n",
       " {'url': 'https://world-weather.info/forecast/turkey/istanbul/august-2024/',\n",
       "  'content': 'Extended weather forecast in Istanbul. Hourly Week 10 days 14 days 30 days Year. Detailed ⚡ Istanbul Weather Forecast for August 2024 - day/night 🌡️ temperatures, precipitations - World-Weather.info.'},\n",
       " {'url': 'https://weather.com/weather/monthly/l/33d1e415eb66f3e1ab35c3add45fccf4512715d329edbd91c806a6957e123b49',\n",
       "  'content': 'Monthly Weather-Istanbul, Türkiye. As of 4:18 am GMT+03:00 ... 26. 90 ° 72 ° 27. 85 ° 71 ° 28. 84 ... Weather in your inbox. Your local forecast, plus daily trivia, stunning photos and our ...'}]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "web_search_tool.invoke(\"What is the weather like in Istanbul?\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### GraphState"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from typing import List\n",
    "\n",
    "from typing_extensions import TypedDict\n",
    "\n",
    "\n",
    "class GraphState(TypedDict):\n",
    "    \"\"\"\n",
    "    Represents the state of our graph.\n",
    "\n",
    "    Attributes:\n",
    "        question: question\n",
    "        generation: LLM generation\n",
    "        documents: list of documents\n",
    "    \"\"\"\n",
    "\n",
    "    question: str\n",
    "    generation: str\n",
    "    documents: List[str]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.schema import Document\n",
    "\n",
    "def retrieve(state):\n",
    "    \"\"\"\n",
    "    Retrieve documents\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): New key added to state, documents, that contains retrieved documents\n",
    "    \"\"\"\n",
    "    print(\"---RETRIEVE---\")\n",
    "    question = state[\"question\"]\n",
    "\n",
    "    # Retrieval\n",
    "    documents = retriever.invoke(question)\n",
    "    return {\"documents\": documents, \"question\": question}\n",
    "\n",
    "def translator(state):\n",
    "    \"\"\"\n",
    "    Translate the question into English if needed.\n",
    "\n",
    "    Args:\n",
    "        state(dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): The question translated to English language.\n",
    "    \"\"\"\n",
    "    print(\"---TRANSLATE THE QUERY TO ENGLISH---\")\n",
    "    question = state[\"question\"]\n",
    "\n",
    "    # Translate\n",
    "    translated_question = question_translater.invoke(question)\n",
    "    return {\"question\": translated_question}\n",
    "\n",
    "\n",
    "def generate(state):\n",
    "    \"\"\"\n",
    "    Generate answer\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): New key added to state, generation, that contains LLM generation\n",
    "    \"\"\"\n",
    "    print(\"---GENERATE---\")\n",
    "    question = state[\"question\"]\n",
    "    documents = state[\"documents\"]\n",
    "\n",
    "    # RAG generation\n",
    "    generation = rag_chain.invoke({\"context\": documents, \"question\": question})\n",
    "    return {\"documents\": documents, \"question\": question, \"generation\": generation}\n",
    "\n",
    "\n",
    "def grade_documents(state):\n",
    "    \"\"\"\n",
    "    Determines whether the retrieved documents are relevant to the question.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): Updates documents key with only filtered relevant documents\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---CHECK DOCUMENT RELEVANCE TO QUESTION---\")\n",
    "    question = state[\"question\"]\n",
    "    documents = state[\"documents\"]\n",
    "\n",
    "    # Score each doc\n",
    "    filtered_docs = []\n",
    "    for d in documents:\n",
    "        score = retrieval_grader.invoke(\n",
    "            {\"question\": question, \"document\": d.page_content}\n",
    "        )\n",
    "        grade = score.binary_score\n",
    "        if grade == \"yes\":\n",
    "            print(\"---GRADE: DOCUMENT RELEVANT---\")\n",
    "            filtered_docs.append(d)\n",
    "        else:\n",
    "            print(\"---GRADE: DOCUMENT NOT RELEVANT---\")\n",
    "            continue\n",
    "    return {\"documents\": filtered_docs, \"question\": question}\n",
    "\n",
    "\n",
    "def transform_query(state):\n",
    "    \"\"\"\n",
    "    Transform the query to produce a better question.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): Updates question key with a re-phrased question\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---TRANSFORM QUERY---\")\n",
    "    question = state[\"question\"]\n",
    "    documents = state[\"documents\"]\n",
    "\n",
    "    # Re-write question\n",
    "    better_question = question_rewriter.invoke({\"question\": question})\n",
    "    return {\"documents\": documents, \"question\": better_question}\n",
    "\n",
    "\n",
    "def web_search(state):\n",
    "    \"\"\"\n",
    "    Web search based on the re-phrased question.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): Updates documents key with appended web results\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---WEB SEARCH---\")\n",
    "    question = state[\"question\"]\n",
    "\n",
    "    # Web search\n",
    "    docs = web_search_tool.invoke({\"query\": question})\n",
    "    web_results = \"\\n\".join([d[\"content\"] for d in docs])\n",
    "    web_results = Document(page_content=web_results)\n",
    "\n",
    "    return {\"documents\": web_results, \"question\": question}\n",
    "\n",
    "def last_response_evaluator(state):\n",
    "    \"\"\"\n",
    "    It evaluates the answer before it is shown to the user.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict):\n",
    "    \"\"\"\n",
    "\n",
    "    question = state[\"question\"]\n",
    "    documents = state[\"documents\"]\n",
    "    final_response = state[\"generation\"]\n",
    "\n",
    "    response_check = answer_evaluator.invoke({\"generation\": final_response})\n",
    "    print(\"---RESPONSE CHECK---\")\n",
    "    if response_check == \"yes\":\n",
    "        return {\"generation\":\"We cannot answer you, please ask another question.\"}\n",
    "    if response_check == \"no\":\n",
    "        return {\"documents\": documents, \"question\": question, \"generation\": final_response}\n",
    "\n",
    "\n",
    "### Edges ###\n",
    "\n",
    "\n",
    "def route_question(state):\n",
    "    \"\"\"\n",
    "    Route question to web search or RAG.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        str: Next node to call\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---ROUTE QUESTION---\")\n",
    "    question = state[\"question\"]\n",
    "    source = question_router.invoke({\"question\": question})\n",
    "    if source.datasource == \"web_search\":\n",
    "        print(\"---ROUTE QUESTION TO WEB SEARCH---\")\n",
    "        return \"web_search\"\n",
    "    elif source.datasource == \"vectorstore\":\n",
    "        print(\"---ROUTE QUESTION TO RAG---\")\n",
    "        return \"vectorstore\"\n",
    "\n",
    "\n",
    "def decide_to_generate(state):\n",
    "    \"\"\"\n",
    "    Determines whether to generate an answer, or re-generate a question.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        str: Binary decision for next node to call\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---ASSESS GRADED DOCUMENTS---\")\n",
    "    state[\"question\"]\n",
    "    filtered_documents = state[\"documents\"]\n",
    "\n",
    "    if not filtered_documents:\n",
    "        # All documents have been filtered check_relevance\n",
    "        # We will re-generate a new query\n",
    "        print(\n",
    "            \"---DECISION: ALL DOCUMENTS ARE NOT RELEVANT TO QUESTION, TRANSFORM QUERY---\"\n",
    "        )\n",
    "        return \"transform_query\"\n",
    "    else:\n",
    "        # We have relevant documents, so generate answer\n",
    "        print(\"---DECISION: GENERATE---\")\n",
    "        return \"generate\"\n",
    "\n",
    "\n",
    "def grade_generation_v_documents_and_question(state):\n",
    "    \"\"\"\n",
    "    Determines whether the generation is grounded in the document and answers question.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        str: Decision for next node to call\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---CHECK HALLUCINATIONS---\")\n",
    "    question = state[\"question\"]\n",
    "    documents = state[\"documents\"]\n",
    "    generation = state[\"generation\"]\n",
    "\n",
    "    score = hallucination_grader.invoke(\n",
    "        {\"documents\": documents, \"generation\": generation}\n",
    "    )\n",
    "    grade = score.binary_score\n",
    "\n",
    "    # Check hallucination\n",
    "    if grade == \"yes\":\n",
    "        print(\"---DECISION: GENERATION IS GROUNDED IN DOCUMENTS---\")\n",
    "        # Check question-answering\n",
    "        print(\"---GRADE GENERATION vs QUESTION---\")\n",
    "        score = answer_grader.invoke({\"question\": question, \"generation\": generation})\n",
    "        grade = score.binary_score\n",
    "        if grade == \"yes\":\n",
    "            print(\"---DECISION: GENERATION ADDRESSES QUESTION---\")\n",
    "            return \"useful\"\n",
    "        else:\n",
    "            print(\"---DECISION: GENERATION DOES NOT ADDRESS QUESTION---\")\n",
    "            return \"not useful\"\n",
    "    else:\n",
    "        pprint(\"---DECISION: GENERATION IS NOT GROUNDED IN DOCUMENTS, RE-TRY---\")\n",
    "        return \"not supported\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build Graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langgraph.graph import END, StateGraph, START\n",
    "\n",
    "workflow = StateGraph(GraphState)\n",
    "\n",
    "# Define the nodes\n",
    "workflow.add_node(\"translate_query\", translator)  # translator\n",
    "workflow.add_node(\"web_search\", web_search)  # web search\n",
    "workflow.add_node(\"retrieve\", retrieve)  # retrieve\n",
    "workflow.add_node(\"grade_documents\", grade_documents)  # grade documents\n",
    "workflow.add_node(\"generate\", generate)  # generate\n",
    "workflow.add_node(\"transform_query\", transform_query)  # transform_query\n",
    "workflow.add_node(\"last_response_evaluator\", last_response_evaluator)  # last_response_evaluator\n",
    "\n",
    "\n",
    "# Build graph\n",
    "workflow.add_conditional_edges(\n",
    "    START,\n",
    "    route_question,\n",
    "    {\n",
    "        \"web_search\": \"web_search\",\n",
    "        \"vectorstore\": \"translate_query\",  # Check if query needs translation\n",
    "    },\n",
    ")\n",
    "workflow.add_edge(\"translate_query\", \"retrieve\")  # Translate query before retrieving\n",
    "workflow.add_edge(\"web_search\", \"generate\")\n",
    "workflow.add_edge(\"retrieve\", \"grade_documents\")\n",
    "workflow.add_conditional_edges(\n",
    "    \"grade_documents\",\n",
    "    decide_to_generate,\n",
    "    {\n",
    "        \"transform_query\": \"transform_query\",\n",
    "        \"generate\": \"generate\",\n",
    "    },\n",
    ")\n",
    "workflow.add_edge(\"transform_query\", \"retrieve\")\n",
    "workflow.add_conditional_edges(\n",
    "    \"generate\",\n",
    "    grade_generation_v_documents_and_question,\n",
    "    {\n",
    "        \"not supported\": \"generate\",\n",
    "        \"useful\": \"last_response_evaluator\",\n",
    "        \"not useful\": \"transform_query\",\n",
    "    },\n",
    ")\n",
    "workflow.add_edge(\"last_response_evaluator\", END)\n",
    "\n",
    "# Compile\n",
    "app = workflow.compile()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### visualize the agent flow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/jpeg": "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",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Draw the graph\n",
    "\n",
    "from IPython.display import Image, display\n",
    "from langchain_core.runnables.graph import MermaidDrawMethod\n",
    "\n",
    "display(\n",
    "    Image(\n",
    "        app.get_graph().draw_mermaid_png(\n",
    "            draw_method=MermaidDrawMethod.API,\n",
    "        )\n",
    "    )\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---ROUTE QUESTION---\n",
      "---ROUTE QUESTION TO RAG---\n",
      "---TRANSLATE THE QUERY TO ENGLISH---\n",
      "---RETRIEVE---\n",
      "---CHECK DOCUMENT RELEVANCE TO QUESTION---\n",
      "---GRADE: DOCUMENT RELEVANT---\n",
      "---GRADE: DOCUMENT NOT RELEVANT---\n",
      "---GRADE: DOCUMENT RELEVANT---\n",
      "---ASSESS GRADED DOCUMENTS---\n",
      "---DECISION: GENERATE---\n",
      "---GENERATE---\n",
      "---CHECK HALLUCINATIONS---\n",
      "---DECISION: GENERATION IS GROUNDED IN DOCUMENTS---\n",
      "---GRADE GENERATION vs QUESTION---\n",
      "---DECISION: GENERATION ADDRESSES QUESTION---\n",
      "---RESPONSE CHECK---\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\"The Hagia Sophia is located in the Sultanahmet district of Istanbul, dominating the Sultanahmet Square and the city's skyline. This iconic structure has a rich history, having been expanded and restored multiple times throughout the centuries. Originally built as a cathedral, it has served various purposes over the years, including being converted into a museum by Mustafa Kemal Atatürk in 1935. Today, it stands as one of the most magnificent tourist attractions in Istanbul, renowned for its stunning design and architecture.\""
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "app.invoke({\"question\":\"Where is the Hagia Sophia located in Istanbul?\"})[\"generation\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## run the agents to get response"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### use web-search tool"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---ROUTE QUESTION---\n",
      "---ROUTE QUESTION TO WEB SEARCH---\n",
      "---WEB SEARCH---\n",
      "---GENERATE---\n",
      "---CHECK HALLUCINATIONS---\n",
      "---DECISION: GENERATION IS GROUNDED IN DOCUMENTS---\n",
      "---GRADE GENERATION vs QUESTION---\n",
      "---DECISION: GENERATION ADDRESSES QUESTION---\n",
      "---RESPONSE CHECK---\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\"As of now, the weather in Istanbul is quite warm and partly cloudy. The current temperature is approximately 31°C (87.8°F), and it feels like 32.2°C (89.9°F) due to humidity and other factors. The humidity level is around 46%, which contributes to the warm feel. \\n\\nThere is a northeast wind blowing at about 16.1 mph (25.9 kph), with gusts reaching up to 17.5 mph (28.2 kph). The atmospheric pressure is measured at 1010 mb, and there has been no precipitation reported, indicating a dry day. Visibility is good, at about 10 kilometers (6 miles).\\n\\nOverall, it’s a sunny day with a few clouds, making it a pleasant time to explore the city. Just remember to stay hydrated and wear sunscreen if you're out and about!\""
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "app.invoke({\"question\":\"What is the weather like in Istanbul now?\"})[\"generation\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### use vectorstore to answer the question"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"import time\n",
    "from pprint import pprint\n",
    "\n",
    "# Run\n",
    "inputs = {\n",
    "    \"question\": \"Who was built the Hagia Sophia?\"\n",
    "    }\n",
    "\n",
    "start = time.time()\n",
    "for output in app.stream(inputs):\n",
    "    for key, value in output.items():\n",
    "        # Node\n",
    "        pprint(f\"Node '{key}':\")\n",
    "        # Optional: print full state at each node\n",
    "        # pprint.pprint(value[\"keys\"], indent=2, width=80, depth=None)\n",
    "    pprint(\"\\n---\\n\")\n",
    "\n",
    "# Final generation\n",
    "pprint(value[\"generation\"])\n",
    "end = time.time()\n",
    "\n",
    "print(f\"Estimated response time: {end-start} seconds.\")\"\"\"\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"import time\n",
    "from pprint import pprint\n",
    "\n",
    "# Run\n",
    "inputs = {\n",
    "    \"question\": \"İstanbul'da gezmek için 3 tane tarihi mekan önerebilir misin? Bu mekanlardan çok kısaca bahset?\"\n",
    "    }\n",
    "\n",
    "start = time.time()\n",
    "for output in app.stream(inputs):\n",
    "    for key, value in output.items():\n",
    "        # Node\n",
    "        pprint(f\"Node '{key}':\")\n",
    "        # Optional: print full state at each node\n",
    "        # pprint.pprint(value[\"keys\"], indent=2, width=80, depth=None)\n",
    "    pprint(\"\\n---\\n\")\n",
    "\n",
    "# Final generation\n",
    "pprint(value[\"generation\"])\n",
    "end = time.time()\n",
    "\n",
    "print(f\"Estimated response time: {end-start} seconds.\")\"\"\"\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"import time\n",
    "from pprint import pprint\n",
    "\n",
    "# Run\n",
    "inputs = {\n",
    "    \"question\": \"Какие текущие цены на общественный транспорт в Стамбул?\"\n",
    "    }\n",
    "\n",
    "start = time.time()\n",
    "for output in app.stream(inputs):\n",
    "    for key, value in output.items():\n",
    "        # Node\n",
    "        pprint(f\"Node '{key}':\")\n",
    "        # Optional: print full state at each node\n",
    "        # pprint.pprint(value[\"keys\"], indent=2, width=80, depth=None)\n",
    "    pprint(\"\\n---\\n\")\n",
    "\n",
    "# Final generation\n",
    "pprint(value[\"generation\"])\n",
    "end = time.time()\n",
    "\n",
    "print(f\"Estimated response time: {end-start} seconds.\")\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---ROUTE QUESTION---\n",
      "---ROUTE QUESTION TO RAG---\n",
      "---TRANSLATE THE QUERY TO ENGLISH---\n",
      "---RETRIEVE---\n",
      "---CHECK DOCUMENT RELEVANCE TO QUESTION---\n",
      "---GRADE: DOCUMENT RELEVANT---\n",
      "---GRADE: DOCUMENT RELEVANT---\n",
      "---GRADE: DOCUMENT NOT RELEVANT---\n",
      "---ASSESS GRADED DOCUMENTS---\n",
      "---DECISION: GENERATE---\n",
      "---GENERATE---\n",
      "---CHECK HALLUCINATIONS---\n",
      "---DECISION: GENERATION IS GROUNDED IN DOCUMENTS---\n",
      "---GRADE GENERATION vs QUESTION---\n",
      "---DECISION: GENERATION ADDRESSES QUESTION---\n",
      "---RESPONSE CHECK---\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\"The Suleymaniye Mosque is one of the most significant architectural masterpieces in Istanbul, dominating the city's skyline from its prominent position on one of the highest hills. It was commissioned by Sultan Suleyman the Magnificent, who ruled during the height of the Ottoman Empire, and was completed in 1557. The mosque was designed by the renowned Ottoman chief architect Mimar Sinan, who is celebrated for his contributions to the architectural landscape of Istanbul, having designed numerous buildings throughout the city.\\n\\nThe Suleymaniye Mosque was conceived as an Islamic response to the grandeur of the Byzantine Hagia Sophia, showcasing the Ottoman Empire's architectural prowess. The mosque features a vast central dome, which is surrounded by tiered domes, creating a stunning visual effect both from the outside and within. The interior of the mosque is particularly impressive, characterized by its spaciousness and the intricate details of its design.\\n\\nMimar Sinan's work on the Suleymaniye Mosque is often regarded as a culmination of his architectural achievements, as he perfected the use of the central dome in mosque design. The mosque not only serves as a place of worship but also as a complex that includes a madrasa (educational institution), a hospital, a public kitchen, and a caravanserai, reflecting the multifaceted role of mosques in Ottoman society.\\n\\nOverall, the Suleymaniye Mosque stands as a testament to the artistic and cultural zenith of the Ottoman Empire, embodying both religious significance and architectural innovation.\""
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "app.invoke({\"question\":\"쉴레이마니예 모스크의 역사에 대한 정보를 제공하세요?\"})[\"generation\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}