generate_dataset.py

import json
import random

conditions = [
    "Hypertension", "Diabetes", "Asthma", "Arthritis", "Depression",
    "Anxiety", "Obesity", "Migraine", "Allergies", "Influenza"
]

medications = [
    "Lisinopril", "Metformin", "Albuterol", "Ibuprofen", "Sertraline",
    "Alprazolam", "Orlistat", "Sumatriptan", "Cetirizine", "Oseltamivir"
]

def generate_question(condition):
    questions = [
        f"What are the symptoms of {condition}?",
        f"How is {condition} typically diagnosed?",
        f"What are the common treatments for {condition}?",
        f"Can you explain the causes of {condition}?",
        f"What lifestyle changes can help manage {condition}?",
        f"Are there any complications associated with {condition}?",
        f"How can {condition} be prevented?",
        f"What's the long-term outlook for someone with {condition}?",
        f"Are there any new treatments being developed for {condition}?",
        f"How does {condition} affect daily life?"
    ]
    return random.choice(questions)

# Function to generate an answer (simplified for this example)
def generate_answer(condition, question):
    return f"Here's some information about {condition} related to your question: '{question}' [Detailed medical explanation would go here.]"

# Function to generate a medication question
def generate_medication_question(medication):
    questions = [
        f"What is {medication} used for?",
        f"What are the common side effects of {medication}?",
        f"How should {medication} be taken?",
        f"Are there any drug interactions with {medication}?",
        f"What should I know before starting {medication}?",
        f"How long does it take for {medication} to start working?",
        f"Can {medication} be taken during pregnancy?",
        f"What should I do if I miss a dose of {medication}?",
        f"Is {medication} habit-forming?",
        f"Are there any alternatives to {medication}?"
    ]
    return random.choice(questions)

def generate_medication_answer(medication, question):
    return f"Regarding {medication} and your question: '{question}' [Detailed medication information would go here.]"

dataset = []
for _ in range(5000):  
    if random.choice([True, False]):
        condition = random.choice(conditions)
        question = generate_question(condition)
        answer = generate_answer(condition, question)
    else:
        medication = random.choice(medications)
        question = generate_medication_question(medication)
        answer = generate_medication_answer(medication, question)
    
    dataset.append({
        "question": question,
        "answer": answer,
        "text": f"Question: {question}\nAnswer: {answer}"
    })

with open("medical_dataset.json", "w") as f:
    json.dump(dataset, f, indent=2)

print("Dataset generated and saved to medical_dataset.json")