fastapi / app.py
Prajith04's picture
Update app.py
94fe369 verified
from fastapi import FastAPI
from pydantic import BaseModel
from sentence_transformers import SentenceTransformer
import chromadb
from fastapi.middleware.cors import CORSMiddleware
import uvicorn
import requests
from itertools import combinations
import sqlite3
import pandas as pd
import os
import time
# Define FastAPI app
app = FastAPI()
origins = [
"http://localhost:5173",
"localhost:5173"
]
app.add_middleware(
CORSMiddleware,
allow_origins=origins,
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"]
)
# Load the model at startup
model = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
client = chromadb.PersistentClient(path='./chromadb')
collection = client.get_or_create_collection(name="symptomsvector")
# Helper function to initialize database and populate from CSV if needed
def init_db():
conn = sqlite3.connect("diseases_symptoms.db")
cursor = conn.cursor()
cursor.execute('''
CREATE TABLE IF NOT EXISTS diseases (
id INTEGER PRIMARY KEY,
name TEXT,
symptoms TEXT,
treatments TEXT
)
''')
conn.commit()
return conn
# Populate database from CSV if it's the first time
if not os.path.exists("diseases_symptoms.db"):
conn = init_db()
df = pd.read_csv("hf://datasets/QuyenAnhDE/Diseases_Symptoms/Diseases_Symptoms.csv")
df['Symptoms'] = df['Symptoms'].str.split(',').apply(lambda x: [s.strip() for s in x])
for _, row in df.iterrows():
symptoms_str = ",".join(row['Symptoms'])
cursor = conn.cursor()
cursor.execute("INSERT INTO diseases (name, symptoms, treatments) VALUES (?, ?, ?)",
(row['Name'], symptoms_str, row.get('Treatments', '')))
conn.commit()
conn.close()
class SymptomQuery(BaseModel):
symptom: str
# Helper function to fetch diseases matching symptoms from SQLite
def fetch_diseases_by_symptoms(matching_symptoms):
conn = sqlite3.connect("diseases_symptoms.db")
cursor = conn.cursor()
disease_list = []
unique_symptoms_list = []
matching_symptom_str = ','.join(matching_symptoms)
# Retrieve matching diseases based on symptoms in SQLite
for row in cursor.execute("SELECT name, symptoms, treatments FROM diseases WHERE symptoms LIKE ?",
(f'%{matching_symptom_str}%',)):
disease_info = {
'Disease': row[0],
'Symptoms': row[1].split(','),
'Treatments': row[2]
}
disease_list.append(disease_info)
# Add symptoms to the unique list, converting to lowercase to avoid duplicates
for symptom in row[1].split(','):
symptom_lower = symptom.strip().lower()
if symptom_lower not in unique_symptoms_list:
unique_symptoms_list.append(symptom_lower)
conn.close()
return disease_list, unique_symptoms_list
@app.post("/find_matching_symptoms")
def find_matching_symptoms(query: SymptomQuery):
symptoms = query.symptom.split(',')
all_results = []
for symptom in symptoms:
symptom = symptom.strip()
query_embedding = model.encode([symptom])
# Perform similarity search in ChromaDB
results = collection.query(
query_embeddings=query_embedding.tolist(),
n_results=3
)
all_results.extend(results['documents'][0])
matching_symptoms = list(dict.fromkeys(all_results))
return {"matching_symptoms": matching_symptoms}
@app.post("/find_disease_list")
def find_disease_list(query: SymptomQuery):
# Normalize and embed each input symptom
selected_symptoms = [symptom.strip().lower() for symptom in query.symptom.split(',')]
all_selected_symptoms.update(selected_symptoms) # Add new symptoms to the set
all_results = []
for symptom in selected_symptoms:
# Generate the embedding for the current symptom
query_embedding = model.encode([symptom])
# Perform similarity search in ChromaDB
results = collection.query(
query_embeddings=query_embedding.tolist(),
n_results=5 # Return top 5 similar symptoms for each input symptom
)
# Aggregate the matching symptoms from the results
all_results.extend(results['documents'][0])
# Remove duplicates while preserving order
matching_symptoms = list(dict.fromkeys(all_results))
conn = sqlite3.connect("diseases_symptoms.db")
cursor = conn.cursor()
disease_list = []
unique_symptoms_set = set()
# Retrieve diseases that contain any of the matching symptoms
for row in cursor.execute("SELECT name, symptoms, treatments FROM diseases"):
disease_name = row[0]
disease_symptoms = [symptom.strip().lower() for symptom in row[1].split(',')] # Normalize database symptoms
treatments = row[2]
# Check if there is any overlap between matching symptoms and the disease symptoms
matched_symptoms = [symptom for symptom in matching_symptoms if symptom in disease_symptoms]
if matched_symptoms: # Include disease if there is at least one matching symptom
disease_info = {
'Disease': disease_name,
'Symptoms': disease_symptoms,
'Treatments': treatments
}
disease_list.append(disease_info)
# Add symptoms not yet selected by the user to unique symptoms list
for symptom in disease_symptoms:
if symptom not in selected_symptoms:
unique_symptoms_set.add(symptom)
conn.close()
# Convert unique symptoms set to a sorted list for consistent output
unique_symptoms_list = sorted(unique_symptoms_set)
return {
"disease_list": disease_list,
"unique_symptoms_list": unique_symptoms_list
}
class SelectedSymptomsQuery(BaseModel):
selected_symptoms: list
# Initialize global list for persistent selected symptoms
all_selected_symptoms = set() # Use a set to avoid duplicates
@app.post("/find_disease")
def find_disease(query: SelectedSymptomsQuery):
# Normalize input symptoms and add them to global list
new_symptoms = [symptom.strip().lower() for symptom in query.selected_symptoms]
all_selected_symptoms.update(new_symptoms) # Add new symptoms to the set
conn = sqlite3.connect("diseases_symptoms.db")
cursor = conn.cursor()
disease_list = []
unique_symptoms_set = set()
# Fetch all diseases and calculate matching symptoms
for row in cursor.execute("SELECT name, symptoms, treatments FROM diseases"):
disease_name = row[0]
disease_symptoms = [symptom.strip().lower() for symptom in row[1].split(',')]
treatments = row[2]
# Find common symptoms between all selected and disease symptoms
matched_symptoms = [symptom for symptom in all_selected_symptoms if symptom in disease_symptoms]
# Check for full match between known symptoms and disease symptoms
if len(matched_symptoms) == len(all_selected_symptoms):
disease_info = {
'Disease': disease_name,
'Symptoms': disease_symptoms,
'Treatments': treatments
}
disease_list.append(disease_info)
# Add symptoms not yet selected by the user to unique symptoms list
for symptom in disease_symptoms:
if symptom not in all_selected_symptoms:
unique_symptoms_set.add(symptom)
conn.close()
# Convert unique symptoms set to a sorted list for consistent output
unique_symptoms_list = sorted(unique_symptoms_set)
return {
"unique_symptoms_list": unique_symptoms_list,
"all_selected_symptoms": list(all_selected_symptoms), # Convert set to list for JSON response
"disease_list": disease_list
}
class DiseaseDetail(BaseModel):
Disease: str
Symptoms: list
Treatments: str
MatchCount: int
@app.post("/pass2llm")
def pass2llm(query: DiseaseDetail):
headers = {
"Authorization": "Bearer 2npJaJjnLBj1RGPcGf0QiyAAJHJ_5qqtw2divkpoAipqN9WLG",
"Ngrok-Version": "2"
}
response = requests.get("https://api.ngrok.com/endpoints", headers=headers)
if response.status_code == 200:
llm_api_response = response.json()
public_url = llm_api_response['endpoints'][0]['public_url']
prompt = f"Here is a list of diseases and their details: {query}. Please generate a summary."
llm_headers = {"Content-Type": "application/json"}
llm_payload = {"model": "llama3", "prompt": prompt, "stream": False}
llm_response = requests.post(f"{public_url}/api/generate", headers=llm_headers, json=llm_payload)
if llm_response.status_code == 200:
llm_response_json = llm_response.json()
return {"message": "Successfully passed to LLM!", "llm_response": llm_response_json.get("response")}
else:
return {"message": "Failed to get response from LLM!", "error": llm_response.text}
else:
return {"message": "Failed to get public URL from Ngrok!", "error": response.text}
@app.post("/trigger-reload")
async def trigger_reload():
global all_selected_symptoms
all_selected_symptoms.clear()
return "cleared"
# To run the FastAPI app with Uvicorn
# if __name__ == "__main__":
# uvicorn.run(app, host="0.0.0.0", port=8000)