{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.agents import create_csv_agent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain.llms import OpenAI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "agent = create_csv_agent(OpenAI(temperature=0), '.inner/xbxb.csv', verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
      "\u001b[32;1m\u001b[1;3mThought: I need to count the rows\n",
      "Action: python_repl_ast\n",
      "Action Input: len(df)\u001b[0m\n",
      "Observation: \u001b[36;1m\u001b[1;3m59867\u001b[0m\n",
      "Thought:\u001b[32;1m\u001b[1;3m I now know the final answer\n",
      "Final Answer: There are 59867 rows.\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'There are 59867 rows.'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"how many rows are there?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
      "\u001b[32;1m\u001b[1;3mThought: I need to find out what values are in the domain释义 column\n",
      "Action: python_repl_ast\n",
      "Action Input: df['domain释义'].unique()\u001b[0m\n",
      "Observation: \u001b[36;1m\u001b[1;3m[nan '歌曲点播' '天气查询' '笑话' '诗词背诵' '故事点播' '音量控制' '电量查询' '屏幕亮度' '绕口令' '十万个为什么'\n",
      " '儿童聊天' '维基百科' '音量调节' '日期查询' '四则运算' '拍照' '知识库' '意图不明确' '闹钟' '休眠']\u001b[0m\n",
      "Thought:\u001b[32;1m\u001b[1;3m I now know the final answer\n",
      "Final Answer: domain释义这一列有nan '歌曲点播' '天气查询' '笑话' '诗词背诵' '故事点播' '音量控制' '电量查询' '屏幕亮度' '绕口令' '十万个为什么' '儿童聊天' '维基百科' '音量调节' '日期查询' '四则运算' '拍照' '知识库' '意图不明确' '闹钟' '休眠'\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\"domain释义这一列有nan '歌曲点播' '天气查询' '笑话' '诗词背诵' '故事点播' '音量控制' '电量查询' '屏幕亮度' '绕口令' '十万个为什么' '儿童聊天' '维基百科' '音量调节' '日期查询' '四则运算' '拍照' '知识库' '意图不明确' '闹钟' '休眠'\""
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"domain释义这一列，有哪些内容\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
      "\u001b[32;1m\u001b[1;3mThought: I need to find out what type of domain释义 is most common\n",
      "Action: python_repl_ast\n",
      "Action Input: df['domain释义'].value_counts()\u001b[0m\n",
      "Observation: \u001b[36;1m\u001b[1;3m儿童聊天      9037\n",
      "意图不明确     4335\n",
      "十万个为什么    3585\n",
      "日期查询      2492\n",
      "诗词背诵      2174\n",
      "知识库       1127\n",
      "天气查询      1125\n",
      "四则运算      1072\n",
      "维基百科       735\n",
      "笑话         531\n",
      "绕口令        250\n",
      "故事点播       245\n",
      "歌曲点播       226\n",
      "闹钟          64\n",
      "音量调节        58\n",
      "屏幕亮度        41\n",
      "电量查询        41\n",
      "音量控制        41\n",
      "拍照          16\n",
      "休眠           5\n",
      "Name: domain释义, dtype: int64\u001b[0m\n",
      "Thought:\u001b[32;1m\u001b[1;3m 儿童聊天类型最多\n",
      "Final Answer: 儿童聊天类型最多\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'儿童聊天类型最多'"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"domain释义这一列，哪种类型最多\")"
   ]
  }
 ],
 "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": "3.10.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}