CatMapperAPI/rag.py at main · ProjectCatMapper/CatMapperAPI · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
# rag_neo4j.py

from sentence_transformers import SentenceTransformer
import chromadb
import openai
from openai import OpenAI


# ------------------- CONFIG -------------------

# Neo4j Config

# OpenAI API key
openai.api_key = api_key

# Chroma settings
CHROMA_PERSIST_DIR = "./chroma"

# ------------------- STEP 1: Extract Data from Neo4j -------------------


# def get_graph_data():
#     driver = GraphDatabase.driver('neo4j://sociomap.rc.asu.edu:7687', auth=(
#                 'u', 'p'))
#     query = """
#     MATCH (n)-[r]->(m)
#     RETURN n, r, m
#     LIMIT 1000
#     """
#     with driver.session() as session:
#         results = session.run(query)
#         records = []
#         for record in results:
#             n = dict(record["n"])
#             r = record["r"].type
#             m = dict(record["m"])
#             n_label = list(record["n"].labels)[0] if record["n"].labels else "Node"
#             m_label = list(record["m"].labels)[0] if record["m"].labels else "Node"
#             sentence = f"{n_label} {n} -[{r}]-> {m_label} {m}"
#             records.append(sentence)
#         return records

# # ------------------- STEP 2: Embed and Store in Vector DB -------------------

# def embed_and_store(texts):
#     # Initialize embedding model
#     embedder = SentenceTransformer("all-MiniLM-L6-v2")
#     embeddings = embedder.encode(texts)

#     # Initialize Chroma
#     chroma_client = chromadb.PersistentClient(path=CHROMA_PERSIST_DIR)
#     collection = chroma_client.get_or_create_collection(name="neo4j_rag")

#     for i, (text, emb) in enumerate(zip(texts, embeddings)):
#         collection.add(documents=[text], embeddings=[emb.tolist()], ids=[str(i)])

#     return embedder, collection

# # ------------------- STEP 3: Retrieve Context -------------------

# def retrieve_context(query, embedder, collection, top_k=3):
#     query_emb = embedder.encode([query])[0]
#     results = collection.query(query_embeddings=[query_emb.tolist()], n_results=top_k)
#     return results["documents"][0]

# # ------------------- STEP 4: Generate Answer -------------------

# def generate_answer(query, context):
#     prompt = f"""
# You are a helpful assistant using the following Neo4j graph knowledge:

# {context}

# User question: {query}
# Answer:
# """
#     response = openai.ChatCompletion.create(
#         model="gpt-4",
#         messages=[{"role": "user", "content": prompt}]
#     )
#     return response['choices'][0]['message']['content']

# # ------------------- MAIN -------------------

# if __name__ == "__main__":
#     print("📦 Extracting data from Neo4j...")
#     texts = get_graph_data()

#     print("🔍 Embedding and storing...")
#     embedder, collection = embed_and_store(texts)

#     while True:
#         query = input("\nAsk a question about the graph (or 'exit'): ")
#         if query.lower() in ["exit", "quit"]:
#             break
#         print("💬 Retrieving context...")
#         context = retrieve_context(query, embedder, collection)
#         print("🧠 Generating answer...")
#         answer = generate_answer(query, context)
#         print(f"\n✅ Answer:\n{answer}")


# Load Chroma collection
client = OpenAI(api_key= api_key)
chroma_client = chromadb.PersistentClient(path=CHROMA_PERSIST_DIR)
collection = chroma_client.get_collection(name="neo4j_rag")

# Load same embedder as before
embedder = SentenceTransformer("all-MiniLM-L6-v2")

# Function to get relevant context
def retrieve_context(query, top_k=3,max_chars=1500):
    query_emb = embedder.encode([query])[0]
    results = collection.query(query_embeddings=[query_emb.tolist()], n_results=top_k)
    docs = results["documents"][0]
    truncated_docs = [doc[:max_chars] for doc in docs]
    return "\n\n".join(truncated_docs)

# Generate answer with OpenAI
def generate_answer(query, context):
    prompt = f"""
You are a helpful assistant using the following Neo4j graph knowledge:

{context}

User question: {query}
Answer:
"""
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": prompt}]
    )
    return response.choices[0].message.content

# --- RAG loop ---
if __name__ == "__main__":
    while True:
        query = input("\nAsk a question (or 'exit'): ")
        if query.lower() in ["exit", "quit"]:
            break
        context = retrieve_context(query)
        answer = generate_answer(query, context)
        print(f"\n✅ Answer:\n{answer}")