# KG: Knowledge Graphs for RAG

### Part 1: Neo4j Environment Setup

step 1: 建立 sandbox 資料庫
連線到 https://neo4j.com/sandbox, 點選左下方的 "Launch the Free Sandbox", 建立 "movie" 專案

step 2: 預備連線資料

NEO4J_URI=bolt://your.neo4j.instance.ip:7687
NEO4J_USERNAME=neo4j
NEO4J_PASSWORD=multisystem-slings-dare

step 3: 確定資料庫連線

setup.py

import dotenv
import os
from neo4j import GraphDatabase, basic_auth

load_status = dotenv.load_dotenv("Neo4j-Sandbox-0619.txt")

url = os.getenv("NEO4J_URI")
username =os.getenv("NEO4J_USERNAME")
password = os.getenv("NEO4J_PASSWORD")

driver = GraphDatabase.driver(
  url,
  auth=basic_auth(username, password))

cypher_query = '''
MATCH (movie:Movie {title:$favorite})<-[:ACTED_IN]-(actor)-[:ACTED_IN]->(rec:Movie)
 RETURN distinct rec.title as title LIMIT 20
'''

with driver.session(database="neo4j") as session:
  results = session.read_transaction(
    lambda tx: tx.run(cypher_query,
                      favorite="The Matrix").data())
  for record in results:
    print(record['title'])

driver.close()

### Part 2: Connect Neo4j Via langchain

example1.py

import dotenv
import os

from langchain_community.graphs import Neo4jGraph

load_status = dotenv.load_dotenv("Neo4j-Sandbox-0619.txt")

url = os.getenv("NEO4J_URI")
username =os.getenv("NEO4J_USERNAME")
password = os.getenv("NEO4J_PASSWORD")

kg = Neo4jGraph(
    url=url, username=username, password=password, database="neo4j"
)

cypher = """
  MATCH (n) 
  RETURN count(n)
  """

result = kg.query(cypher)
print(result)

print(kg.schema)

example2.py

import dotenv
import os

from langchain_community.graphs import Neo4jGraph
from langchain.chains import GraphCypherQAChain
from langchain_openai import ChatOpenAI

load_status = dotenv.load_dotenv("Neo4j-Sandbox-0619.txt")

url = os.getenv("NEO4J_URI")
username =os.getenv("NEO4J_USERNAME")
password = os.getenv("NEO4J_PASSWORD")

kg = Neo4jGraph(
    url=url, username=username, password=password, database="neo4j"
)

chain = GraphCypherQAChain.from_llm(
    # default top_k = 10
    ChatOpenAI(temperature=0), graph=kg, verbose=True

    # limit top_k to 2
    # ChatOpenAI(temperature=0), graph=kg, verbose=True, top_k=2  
)

chain.invoke({"query": "Who played in Top Gun?"})

example3.py

import dotenv
import os

from langchain_community.graphs import Neo4jGraph
from langchain.chains import GraphCypherQAChain
from langchain_openai import ChatOpenAI
from langchain_core.prompts.prompt import PromptTemplate

load_status = dotenv.load_dotenv("Neo4j-Sandbox-0619.txt")

url = os.getenv("NEO4J_URI")
username =os.getenv("NEO4J_USERNAME")
password = os.getenv("NEO4J_PASSWORD")

kg = Neo4jGraph(
    url=url, username=username, password=password, database="neo4j"
)

CYPHER_GENERATION_TEMPLATE = """Task:Generate Cypher statement to query a graph database.
Instructions:
Use only the provided relationship types and properties in the schema.
Do not use any other relationship types or properties that are not provided.
Schema:
{schema}
Note: Do not include any explanations or apologies in your responses.
Do not respond to any questions that might ask anything else than for you to construct a Cypher statement.
Do not include any text except the generated Cypher statement.
Examples: Here are a few examples of generated Cypher statements for particular questions:
# How many people played in Top Gun?
MATCH (m:Movie {{name:"Top Gun"}})<-[:ACTED_IN]-()
RETURN count(*) AS numberOfActors

The question is:
{question}"""

CYPHER_GENERATION_PROMPT = PromptTemplate(
    input_variables=["schema", "question"], template=CYPHER_GENERATION_TEMPLATE
)

chain = GraphCypherQAChain.from_llm(
    llm=ChatOpenAI(temperature=0, model="gpt-3.5-turbo"),
    graph=kg,
    verbose=True,
    cypher_prompt=CYPHER_GENERATION_PROMPT,  
)

chain.invoke({"query": "How many people played in Top Gun?"})
chain.invoke({"query": "Who played in Top Gun?"})

### Part 3: Preparing Text Data for RAG

example4.py

import dotenv
import os

from langchain_community.graphs import Neo4jGraph
from langchain.chains import GraphCypherQAChain
from langchain_openai import ChatOpenAI
from langchain_core.prompts.prompt import PromptTemplate

load_status = dotenv.load_dotenv("Neo4j-Sandbox-0619_ex4.txt")

url = os.getenv("NEO4J_URI")
username =os.getenv("NEO4J_USERNAME")
password = os.getenv("NEO4J_PASSWORD")

# Note the code below is unique to this course environment, and not a 
# standard part of Neo4j's integration with OpenAI. Remove if running 
# in your own environment.
OPENAI_ENDPOINT = os.getenv('OPENAI_BASE_URL') + '/embeddings'
OPENAI_API_KEY = os.getenv('OPENAI_API_KEY')   # This one get from environment variable

kg = Neo4jGraph(
    url=url, username=username, password=password, database="neo4j"
)

# Create a vector index
kg.query("""
  CREATE VECTOR INDEX movie_tagline_embeddings IF NOT EXISTS
  FOR (m:Movie) ON (m.taglineEmbedding) 
  OPTIONS { indexConfig: {
    `vector.dimensions`: 1536,
    `vector.similarity_function`: 'cosine'
  }}"""
)

r_vi=kg.query("""
  SHOW VECTOR INDEXES
  """
)

for index in r_vi:
    print(index)

# Populate the vector index
# - Calculate vector representation for each movie tagline using OpenAI
# - Add vector to the Movie node as taglineEmbedding property
kg.query("""
    MATCH (movie:Movie) WHERE movie.tagline IS NOT NULL
    WITH movie, genai.vector.encode(
        movie.tagline, 
        "OpenAI", 
        {
          token: $openAiApiKey,
          endpoint: $openAiEndpoint
        }) AS vector
    CALL db.create.setNodeVectorProperty(movie, "taglineEmbedding", vector)
    """, 
    params={"openAiApiKey":OPENAI_API_KEY, "openAiEndpoint": OPENAI_ENDPOINT} )

result = kg.query("""
    MATCH (m:Movie) 
    WHERE m.tagline IS NOT NULL
    RETURN m.tagline, m.taglineEmbedding
    LIMIT 1
    """
)

# print(result[0]['m.tagline'])
# print(result[0]['m.taglineEmbedding'][:10])
# print(result[0]['m.taglineEmbedding'])

# Similarity search
# - Calculate embedding for question
# - Identify matching movies based on similarity of question and taglineEmbedding vectors

question = "What movies are about love?"

result2=kg.query("""
    WITH genai.vector.encode(
        $question, 
        "OpenAI", 
        {
          token: $openAiApiKey,
          endpoint: $openAiEndpoint
        }) AS question_embedding
    CALL db.index.vector.queryNodes(
        'movie_tagline_embeddings', 
        $top_k, 
        question_embedding
        ) YIELD node AS movie, score
    RETURN movie.title, movie.tagline, score
    """, 
    params={"openAiApiKey":OPENAI_API_KEY,
            "openAiEndpoint": OPENAI_ENDPOINT,
            "question": question,
            "top_k": 5
            })

# print(f"There are {len(result2)} movies")
# i=len(result2)
# for i in range(len(result2)):
#     print(f"===== Movie # {i+1} =====")
#     print(result2[i]['movie.tagline'])
#     print(result2[i]['movie.title'])
#     print(result2[i]['score'])

---
參考資料
1. DeepLearning.AI: Knowledge Graphs for RAG
https://learn.deeplearning.ai/courses/knowledge-graphs-rag/lesson/1/introduction
2. LangChain Documentation
https://python.langchain.com/v0.2/docs/integrations/graphs/neo4j_cypher/
2.