chat_with_csv.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import necessary libraries\n",
    "import openai\n",
    "import pandas as pd\n",
    "from langchain.agents import create_csv_agent\n",
    "from langchain.llms import AzureOpenAI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load the csv file as Pandas dataframe.\n",
    "df = pd.read_csv('./data/Sales_Performance_Report_DQLab_Store.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 5499 entries, 0 to 5498\n",
      "Data columns (total 10 columns):\n",
      " #   Column                Non-Null Count  Dtype  \n",
      "---  ------                --------------  -----  \n",
      " 0   order_id              5499 non-null   int64  \n",
      " 1   order_status          5499 non-null   object \n",
      " 2   customer              5499 non-null   object \n",
      " 3   order_date            5499 non-null   object \n",
      " 4   order_quantity        5499 non-null   int64  \n",
      " 5   sales                 5499 non-null   int64  \n",
      " 6   discount              5499 non-null   float64\n",
      " 7   discount_value        5499 non-null   int64  \n",
      " 8   product_category      5499 non-null   object \n",
      " 9   product_sub_category  5499 non-null   object \n",
      "dtypes: float64(1), int64(4), object(5)\n",
      "memory usage: 429.7+ KB\n"
     ]
    }
   ],
   "source": [
    "# Display the dataframe information.\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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_DEPLOYMENT_NAME = \"PLEASE_ENTER_YOUR_OWNED_AOAI_TEXT_MODEL_NAME\"\n",
    "MODEL_NAME = \"text-davinci-003\"\n",
    "openai.api_type = \"azure\"\n",
    "openai.api_base = \"https://PLESAE_ENTER_YOUR_OWNED_AOAI_RESOURCE_NAME.openai.azure.com/\"\n",
    "openai.api_version = \"2022-12-01\"\n",
    "openai.api_key = \"PLEASE_ENTER_YOUR_OWNED_AOAI_SERVICE_KEY\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initiate a connection to the LLM from Azure OpenAI Service via LangChain.\n",
    "llm = AzureOpenAI(\n",
    "    openai_api_key=OPENAI_API_KEY,\n",
    "    deployment_name=OPENAI_DEPLOYMENT_NAME,\n",
    "    model_name=MODEL_NAME, \n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "The Azure OpenAI service is a cloud-based service that allows data scientists, developers, and researchers to access the OpenAI platform. It provides users with access to cutting-edge AI technologies, such as machine learning, natural language processing, and computer vision. It also allows users to take advantage of OpenAI’s vast library of pre-trained models, allowing them to quickly create their own AI applications.\n"
     ]
    }
   ],
   "source": [
    "# Make a simple prompt to validate the connection with LLM from Azure OpenAI Service.\n",
    "print(llm(\"What is the Azure OpenAI Service?\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Initiate a LangChain's CSV Agent with the LLM from Azure OpenAI Service to interact with the CSV file.\n",
    "agent = create_csv_agent(AzureOpenAI(openai_api_key=OPENAI_API_KEY, deployment_name=OPENAI_DEPLOYMENT_NAME, model_name=MODEL_NAME), \n",
    "                         './data/Sales_Performance_Report_DQLab_Store.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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}\\n\\nBegin!\\nQuestion: {input}\\n{agent_scratchpad}'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Display the CSV Agent's default prompt template.\n",
    "agent.agent.llm_chain.prompt.template"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'5499 rows of data'"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"How many rows of data do you have?\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'The years in this dataset are 2010, 2012, 2011, and 2009.'"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"Please list the years in this dataset.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'The total sales in 2011 was 4568206579 and the total sales in 2012 was 5071954389, which is a 11.03% increase.'"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"Please tell me the total sales in 2011. Please tell me the total sales in 2012. And calculate the percentage of the increase from 2011 to 2012.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'The total sales in 2013 is estimated to be 9640160968.'"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"Based on the sales trend of 2011 & 2012, please help to forecast total sales in 2013.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<string>:1: FutureWarning: The default value of numeric_only in DataFrameGroupBy.sum is deprecated. In a future version, numeric_only will default to False. Either specify numeric_only or select only columns which should be valid for the function.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "'The customer with the highest total amount of purchases in 2012 is Craig Carreira and the total amount of purchases is 82686420.'"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"What is the customer's name with the highest total amount of purchases in 2012? And tell me the total amount of purchases.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Office Supplies'"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "agent.run(\"What was the top sale product category in 2012?\")"
   ]
  },
  {
   "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"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}