app.py

import streamlit as st
import pandas as pd
import os
from dotenv import load_dotenv
from llama_index.core import Settings, VectorStoreIndex, SimpleDirectoryReader
from llama_index.core.readers.base import BaseReader
from llama_index.readers.file.paged_csv.base import PagedCSVReader
from llama_index.embeddings.openai import OpenAIEmbedding
from llama_index.llms.openai import OpenAI
from llama_index.vector_stores.faiss import FaissVectorStore
from llama_index.core.ingestion import IngestionPipeline
from langchain_community.document_loaders.csv_loader import CSVLoader
from langchain_community.vectorstores import FAISS as LangChainFAISS
from langchain.chains import create_retrieval_chain
from langchain.chains.combine_documents import create_stuff_documents_chain
from langchain_core.prompts import ChatPromptTemplate
from langchain_openai import OpenAIEmbeddings, ChatOpenAI
import faiss

# Load environment variables
os.environ["OPENAI_API_KEY"] = os.getenv("OPENAI_API_KEY")

# Global settings for LlamaIndex
EMBED_DIMENSION = 512
Settings.llm = OpenAI(model="gpt-3.5-turbo")
Settings.embed_model = OpenAIEmbedding(model="text-embedding-3-small", dimensions=EMBED_DIMENSION)

# Streamlit app
st.title("Chat w CSV Files - LangChain Vs LlamaIndex ")

# File uploader
uploaded_file = st.file_uploader("Upload a CSV file", type=["csv"])
if uploaded_file:
    # Save the uploaded file temporarily
    temp_file_path = f"temp_{uploaded_file.name}"
    with open(temp_file_path, "wb") as temp_file:
        temp_file.write(uploaded_file.getbuffer())

    # Read and preview CSV data
    data = pd.read_csv(temp_file_path)
    st.write("Preview of uploaded data:")
    st.dataframe(data)

    # Tabs
    tab1, tab2 = st.tabs(["Chat w CSV using LangChain", "Chat w CSV using LlamaIndex"])

    # LangChain Tab
    with tab1:
        st.subheader("LangChain Query")
        loader = CSVLoader(file_path=temp_file_path)
        docs = loader.load_and_split()

        # Preview the first document
        if docs:
            st.write("Preview of a document chunk (LangChain):")
            st.text(docs[0].page_content)

        # LangChain FAISS VectorStore
        langchain_index = faiss.IndexFlatL2(EMBED_DIMENSION)
        langchain_vector_store = LangChainFAISS(
            embedding_function=OpenAIEmbeddings(),
            index=langchain_index,
        )
        langchain_vector_store.add_documents(docs)

        # LangChain Retrieval Chain
        retriever = langchain_vector_store.as_retriever()
        system_prompt = (
            "You are an assistant for question-answering tasks. "
            "Use the following pieces of retrieved context to answer "
            "the question. If you don't know the answer, say that you "
            "don't know. Use three sentences maximum and keep the "
            "answer concise.\n\n{context}"
        )
        prompt = ChatPromptTemplate.from_messages(
            [("system", system_prompt), ("human", "{input}")]
        )
        question_answer_chain = create_stuff_documents_chain(ChatOpenAI(), prompt)
        langchain_rag_chain = create_retrieval_chain(retriever, question_answer_chain)

        # Query input for LangChain
        query = st.text_input("Ask a question about your data (LangChain):")
        if query:
            answer = langchain_rag_chain.invoke({"input": query})
            st.write(f"Answer: {answer['answer']}")

    # LlamaIndex Tab
    with tab2:
        st.subheader("LlamaIndex Query")
        csv_reader = PagedCSVReader()
        reader = SimpleDirectoryReader(
            input_files=[temp_file_path],
            file_extractor={".csv": csv_reader},
        )
        docs = reader.load_data()

        # Preview the first document
        if docs:
            st.write("Preview of a document chunk (LlamaIndex):")
            st.text(docs[0].text)

        # Initialize FAISS Vector Store
        llama_faiss_index = faiss.IndexFlatL2(EMBED_DIMENSION)
        llama_vector_store = FaissVectorStore(faiss_index=llama_faiss_index)

        # Create the ingestion pipeline and process the data
        pipeline = IngestionPipeline(vector_store=llama_vector_store, documents=docs)
        nodes = pipeline.run()

        # Create a query engine
        llama_index = VectorStoreIndex(nodes)
        query_engine = llama_index.as_query_engine(similarity_top_k=3)

        # Query input for LlamaIndex
        query = st.text_input("Ask a question about your data (LlamaIndex):")
        if query:
            response = query_engine.query(query)
            st.write(f"Answer: {response.response}")

    # Cleanup temporary file
    os.remove(temp_file_path)