nbs/23-05-26_osmnx-tool.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "9ec30f55-baa1-486c-9277-8e80e262edeb",
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append(\"..\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "0321b936-0620-44d9-ab94-5afe2cc02baa",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "import osmnx as ox\n",
    "from osmnx import utils_graph\n",
    "import geopandas as gpd\n",
    "\n",
    "from langchain.chat_models import ChatOpenAI\n",
    "from langchain.tools import BaseTool, DuckDuckGoSearchRun\n",
    "from langchain.agents import Tool\n",
    "\n",
    "from tools.mercantile_tool import MercantileTool\n",
    "from tools.geopy.geocode import GeopyGeocodeTool\n",
    "from tools.geopy.distance import GeopyDistanceTool\n",
    "from agents.l4m_agent import base_agent\n",
    "\n",
    "from getpass import getpass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "8152a613-7b88-4bec-b0a7-8b9256f3c763",
   "metadata": {},
   "outputs": [
    {
     "name": "stdin",
     "output_type": "stream",
     "text": [
      " ········\n"
     ]
    }
   ],
   "source": [
    "OPENAI_API_KEY = getpass()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5eaa8573-86f2-4483-bcc9-7b9c82916f16",
   "metadata": {},
   "outputs": [],
   "source": [
    "os.environ[\"OPENAI_API_KEY\"] = OPENAI_API_KEY"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "4c3fbc82-af1b-40a1-8edc-248c62f93fcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "from typing import Type, Dict\n",
    "\n",
    "from pydantic import BaseModel, Field\n",
    "from langchain.tools import BaseTool\n",
    "\n",
    "class PlaceWithTags(BaseModel):\n",
    "    \"Name of a place and tags in OSM.\"\n",
    "\n",
    "    place: str = Field(..., description=\"name of a place\")\n",
    "    tags: Dict[str, str] = Field(..., description=\"open street maps tags\")\n",
    "\n",
    "\n",
    "class OSMnxGeometryTool(BaseTool):\n",
    "    \"\"\"Custom tool to query geometries from OSM.\"\"\"\n",
    "\n",
    "    name: str = \"geometry\"\n",
    "    args_schema: Type[BaseModel] = PlaceWithTags\n",
    "    description: str = \"Use this tool to get geometry of different features of a place like building footprints, parks, lakes, hospitals, schools etc. \\\n",
    "    Pass the name of the place & relevant tags of Open Street Map as args.\"\n",
    "    return_direct = True\n",
    "\n",
    "    def _run(self, place: str, tags: Dict[str, str]) -> gpd.GeoDataFrame:\n",
    "        gdf = ox.geometries_from_place(place, tags)\n",
    "        gdf = gdf[gdf[\"geometry\"].type.isin({\"Polygon\", \"MultiPolygon\"})]\n",
    "        gdf = gdf[[\"name\", \"geometry\"]].reset_index(drop=True).head(20)\n",
    "        return gdf\n",
    "\n",
    "    def _arun(self, place: str):\n",
    "        raise NotImplementedError\n",
    "\n",
    "class PlaceWithNetworktype(BaseModel):\n",
    "    \"Name of a place on the map\"\n",
    "    place: str = Field(..., description=\"name of a place on the map\")\n",
    "    network_type: str = Field(..., description=\"network type: one of walk, bike, drive or all\")\n",
    "\n",
    "class OSMnxNetworkTool(BaseTool):\n",
    "    \"\"\"Custom tool to query road networks from OSM.\"\"\"\n",
    "\n",
    "    name: str = \"network\"\n",
    "    args_schema: Type[BaseModel] = PlaceWithNetworktype\n",
    "    description: str = \"Use this tool to get road network of a place. \\\n",
    "    Pass the name of the place & type of road network i.e walk, bike, drive or all.\"\n",
    "    return_direct = True\n",
    "\n",
    "    def _run(self, place: str, network_type: str) -> gpd.GeoDataFrame:\n",
    "        G = ox.graph_from_place(place, network_type=network_type, simplify=True)\n",
    "        network = utils_graph.graph_to_gdfs(G, nodes=False) \n",
    "        network = network[[\"name\", \"geometry\"]].reset_index(drop=True).head(20)\n",
    "        return network\n",
    "        \n",
    "    def _arun(self, place: str):\n",
    "        raise NotImplementedError"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "9893d0e0-dd11-4fa2-82f7-da6a37ba83a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "geometry_tool = OSMnxGeometryTool()\n",
    "network_tool = OSMnxNetworkTool()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "153d4abf-4828-4ad3-970c-db47f0772547",
   "metadata": {},
   "outputs": [],
   "source": [
    "# pick a LLM\n",
    "llm = ChatOpenAI(\n",
    "    temperature=0,\n",
    "    openai_api_key=OPENAI_API_KEY,\n",
    "    model_name=\"gpt-3.5-turbo\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "473bf6d6-3489-42a3-ac12-6768b9c4f03b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# define a set of tools the agent has access to for queries\n",
    "duckduckgo_tool = Tool(\n",
    "    name=\"DuckDuckGo\",\n",
    "    description=\"Use this tool to answer questions about current events and places. \\\n",
    "    Please ask targeted questions.\",\n",
    "    func=DuckDuckGoSearchRun().run\n",
    ")\n",
    "\n",
    "geocode_tool = GeopyGeocodeTool()\n",
    "distance_tool = GeopyDistanceTool()\n",
    "mercantile_tool = MercantileTool()\n",
    "\n",
    "tools = [duckduckgo_tool, network_tool, geocode_tool, distance_tool, mercantile_tool, geometry_tool]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ee95931c-b68b-4317-8c1d-227321e0c526",
   "metadata": {},
   "outputs": [],
   "source": [
    "agent = base_agent(llm, tools, name=\"structured-chat-zero-shot-react-description\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3aaf6d47-157f-44e1-9aef-dda05d79c82a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
      "\u001b[32;1m\u001b[1;3mAction:\n",
      "```\n",
      "{\n",
      "  \"action\": \"network\",\n",
      "  \"action_input\": {\n",
      "    \"place\": \"Bangalore, India\",\n",
      "    \"network_type\": \"bike\"\n",
      "  }\n",
      "}\n",
      "``` \n",
      "\u001b[0m\n",
      "Observation: \u001b[33;1m\u001b[1;3m                       name                                           geometry\n",
      "0             2nd Main Road  LINESTRING (77.59872 12.91054, 77.59848 12.91272)\n",
      "1             2nd Main Road  LINESTRING (77.59872 12.91054, 77.59873 12.91049)\n",
      "2            9th Cross Road  LINESTRING (77.59872 12.91054, 77.59899 12.91056)\n",
      "3            7th Cross Road  LINESTRING (77.62408 12.93497, 77.62371 12.93526)\n",
      "4      Mahayogi Vemana Road  LINESTRING (77.62408 12.93497, 77.62437 12.93524)\n",
      "5   3rd Main Ashwini Layout  LINESTRING (77.62917 12.93849, 77.62887 12.93935)\n",
      "6           Inner Ring Road  LINESTRING (77.62917 12.93849, 77.62938 12.938...\n",
      "7            Yelahanka Road  LINESTRING (77.59420 13.09518, 77.59407 13.09533)\n",
      "8                       NaN  LINESTRING (77.59420 13.09518, 77.59427 13.09530)\n",
      "9                       NaN  LINESTRING (77.58396 12.99330, 77.58370 12.99343)\n",
      "10        Kumara Krupa Road  LINESTRING (77.58396 12.99330, 77.58405 12.993...\n",
      "11        Kumara Krupa Road  LINESTRING (77.58396 12.99330, 77.58384 12.993...\n",
      "12                      NaN  LINESTRING (77.57933 12.98596, 77.57841 12.98659)\n",
      "13        Kumara Krupa Road  LINESTRING (77.57933 12.98596, 77.57951 12.98620)\n",
      "14        Kumara Krupa Road  LINESTRING (77.57933 12.98596, 77.57919 12.98580)\n",
      "15                Bazaar st  LINESTRING (77.62001 12.97471, 77.62009 12.97471)\n",
      "16           Bhaskaran Road  LINESTRING (77.62001 12.97471, 77.62001 12.97475)\n",
      "17            Halasuru Road  LINESTRING (77.62001 12.97471, 77.61913 12.974...\n",
      "18           Bhaskaran Road  LINESTRING (77.62023 12.97568, 77.62025 12.97576)\n",
      "19    Gangadhar Chetty Road  LINESTRING (77.62023 12.97568, 77.62010 12.97589)\u001b[0m\n",
      "\u001b[32;1m\u001b[1;3m\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    }
   ],
   "source": [
    "r = agent(\"Find all bike roads in Bangalore, India\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "97b02be6-9bc6-4be3-aa76-313c5ad82f9a",
   "metadata": {},
   "outputs": [],
   "source": [
    "r[\"output\"].to_file(\"roads.geojson\", driver=\"GeoJSON\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fcf0c1d1-9005-4163-9169-4684a3f7fe27",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "r[\"output\"].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "94f4e4a8-99ff-49ee-add4-47b5f95c3842",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}