app.py

from dotenv import load_dotenv
import os
from langchain_community.graphs import Neo4jGraph


from langchain_community.graphs import Neo4jGraph
from langchain_openai import ChatOpenAI

load_dotenv()

AURA_INSTANCENAME = os.environ["AURA_INSTANCENAME"]
NEO4J_URI = os.environ["NEO4J_URI"]
NEO4J_USERNAME = os.environ["NEO4J_USERNAME"]
NEO4J_PASSWORD = os.environ["NEO4J_PASSWORD"]
NEO4J_DATABASE = os.environ["NEO4J_DATABASE"]
AUTH = (NEO4J_USERNAME, NEO4J_PASSWORD)

OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
OPENAI_ENDPOINT = os.getenv("OPENAI_ENDPOINT")

chat = ChatOpenAI(api_key=OPENAI_API_KEY)


kg = Neo4jGraph(
    url=NEO4J_URI,
    username=NEO4J_USERNAME,
    password=NEO4J_PASSWORD,
    database=NEO4J_DATABASE,
)

# kg.query(
#     """
#     CREATE VECTOR INDEX health_providers_embeddings IF NOT EXISTS
#     FOR (hp:HealthcareProvider) ON (hp.comprehensiveEmbedding)
#     OPTIONS {
#       indexConfig: {
#         `vector.dimensions`: 1536,
#         `vector.similarity_function`: 'cosine'
#       }
#     }
#     """
# )

# # test to see if the index was created
# res = kg.query(
#     """
#   SHOW VECTOR INDEXES
#   """
# )
# print(res)

# kg.query(
#     """
#     MATCH (hp:HealthcareProvider)-[:TREATS]->(p:Patient)
#     WHERE hp.bio IS NOT NULL
#     WITH hp, genai.vector.encode(
#         hp.bio,
#         "OpenAI",
#         {
#           token: $openAiApiKey,
#           endpoint: $openAiEndpoint
#         }) AS vector
#     WITH hp, vector
#     WHERE vector IS NOT NULL
#     CALL db.create.setNodeVectorProperty(hp, "comprehensiveEmbedding", vector)
#     """,
#     params={
#         "openAiApiKey": OPENAI_API_KEY,
#         "openAiEndpoint": OPENAI_ENDPOINT,
#     },
# )

# result = kg.query(
#     """
#     MATCH (hp:HealthcareProvider)
#     WHERE hp.bio IS NOT NULL
#     RETURN hp.bio, hp.name, hp.comprehensiveEmbedding
#     LIMIT 5
#     """
# )
# # loop through the results
# for record in result:
#     print(f" bio: {record["hp.bio"]}, name: {record["hp.name"]}")

# == Queerying the graph for a healthcare provider
question = "give me a list of healthcare providers in the area of dermatology"

# # Execute the query
result = kg.query(
    """
    WITH genai.vector.encode(
        $question,
        "OpenAI",
        {
          token: $openAiApiKey,
          endpoint: $openAiEndpoint
        }) AS question_embedding
    CALL db.index.vector.queryNodes(
        'health_providers_embeddings',
        $top_k,
        question_embedding
        ) YIELD node AS healthcare_provider, score
    RETURN healthcare_provider.name, healthcare_provider.bio, score
    """,
    params={
        "openAiApiKey": OPENAI_API_KEY,
        "openAiEndpoint": OPENAI_ENDPOINT,
        "question": question,
        "top_k": 3,
    },
)

# # Print the encoded question vector for debugging
# print("Encoded question vector:", result)

# # Print the result
for record in result:
    print(f"Name: {record['healthcare_provider.name']}")
    print(f"Bio: {record['healthcare_provider.bio']}")
    # print(f"Specialization: {record['healthcare_provider.specialization']}")
    # print(f"Location: {record['healthcare_provider.location']}")
    print(f"Score: {record['score']}")
    print("---")