experiment.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import necessary libraries\n",
    "import openai\n",
    "from langchain.llms import AzureOpenAI\n",
    "from langchain.agents import create_pandas_dataframe_agent\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Configure the baseline configuration of the OpenAI library for Azure OpenAI Service.\n",
    "OPENAI_API_KEY = \"PLEASE_ENTER_YOUR_OWNED_AOAI_SERVICE_KEY\"\n",
    "OPENAI_API_BASE = \"https://PLESAE_ENTER_YOUR_OWNED_AOAI_RESOURCE_NAME.openai.azure.com/\"\n",
    "OPENAI_DEPLOYMENT_NAME = \"PLEASE_ENTER_YOUR_OWNED_AOAI_GPT_MODEL_NAME\"\n",
    "OPENAI_MODEL_NAME = \"text-davinci-003\"\n",
    "OPENAI_API_VERSION = \"2023-05-15\"\n",
    "OPENAI_API_TYPE = \"azure\"\n",
    "openai.api_key = OPENAI_API_KEY\n",
    "openai.api_base = OPENAI_API_BASE\n",
    "openai.api_version = OPENAI_API_VERSION\n",
    "openai.api_type = OPENAI_API_TYPE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the csv file as Pandas dataframe.\n",
    "filename = \"./data/dataset_telco_customer_churn.csv\"\n",
    "df = pd.read_csv(filename)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initiate a connection to the LLM from Azure OpenAI Service via LangChain.\n",
    "llm = AzureOpenAI(openai_api_key=OPENAI_API_KEY, \n",
    "                    deployment_name=OPENAI_DEPLOYMENT_NAME, \n",
    "                    model_name=OPENAI_MODEL_NAME, \n",
    "                    openai_api_version=OPENAI_API_VERSION,\n",
    "                    openai_api_type=OPENAI_API_TYPE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initiate a LangChain's Pandas Agent with the LLM from Azure OpenAI Service to interact with the dataframe.\n",
    "agent = create_pandas_dataframe_agent(llm, df, verbose=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'\\nYou are working with a pandas dataframe in Python. The name of the dataframe is `df`.\\nYou should use the tools below to answer the question posed of you:\\n\\npython_repl_ast: A Python shell. Use this to execute python commands. Input should be a valid python command. When using this tool, sometimes output is abbreviated - make sure it does not look abbreviated before using it in your answer.\\n\\nUse the following format:\\n\\nQuestion: the input question you must answer\\nThought: you should always think about what to do\\nAction: the action to take, should be one of [python_repl_ast]\\nAction Input: the input to the action\\nObservation: the result of the action\\n... (this Thought/Action/Action Input/Observation can repeat N times)\\nThought: I now know the final answer\\nFinal Answer: the final answer to the original input question\\n\\n\\nThis is the result of `print(df.head())`:\\n{df_head}\\n\\nBegin!\\nQuestion: {input}\\n{agent_scratchpad}'"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Display the Pandas Agent's default prompt template.\n",
    "agent.agent.llm_chain.prompt.template"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the preinstruction for the query.\n",
    "instruction = \"\"\"Please say \"Bello!\" before reply to the query. You must need to use matplotlib library to create a any chart.\n",
    "\n",
    "If the query requires creating a bar chart, please save the chart as \"./chart_image/chart.png\" and display it. Please also say \"Here is the chart:\" before displaying the chart.\n",
    "\n",
    "Here is the query: \"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the query.\n",
    "query = \"Please tell me the distribution of churn by total count number of gender in stacked bar chart.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make up the prompt.\n",
    "prompt = instruction + query"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Please say \"Bello!\" before reply to the query. You must need to use matplotlib library to create a any chart.\n",
      "\n",
      "If the query requires creating a bar chart, please save the chart as \"./chart_image/chart.png\" and display it. Please also say \"Here is the chart:\" before displaying the chart.\n",
      "\n",
      "Here is the query: Please tell me the distribution of churn by total count number of gender in stacked bar chart.\n"
     ]
    }
   ],
   "source": [
    "# Display the prompt.\n",
    "print(prompt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
      "\u001b[32;1m\u001b[1;3m\n",
      "Thought: To answer the question, I need to first get the count of churn by gender, then plot the data as a bar chart.\n",
      "Action: python_repl_ast\n",
      "Action Input: df.groupby(['gender','Churn'])['Churn'].count().unstack().plot(kind='bar',stacked=True)\u001b[0m\n",
      "Observation: \u001b[36;1m\u001b[1;3mAxes(0.125,0.11;0.775x0.77)\u001b[0m\n",
      "Thought:\u001b[32;1m\u001b[1;3m I need to save the chart as an image\n",
      "Action: python_repl_ast\n",
      "Action Input: plt.savefig(\"./chart_image/chart.png\")\u001b[0m\n",
      "Observation: \u001b[36;1m\u001b[1;3m\u001b[0m\n",
      "Thought:\u001b[32;1m\u001b[1;3m I now know the final answer\n",
      "Final Answer: Here is the chart: (./chart_image/chart.png)\u001b[0m\n",
      "\n",
      "\u001b[1m> Finished chain.\u001b[0m\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'Here is the chart: (./chart_image/chart.png)'"
      ]
     },
     "execution_count": 189,
     "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": [
    "# Run the agent to interact with the dataframe.\n",
    "agent.run(prompt)"
   ]
  }
 ],
 "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
}