Team 17/final

.gitignore

@@ -0,0 +1,23 @@
+# Git ignore for Freakquency Streamlit app
+
+# Python caches and bytecode
+__pycache__/
+*.py[cod]
+
+# Virtual environments
+.venv/
+venv/
+env/
+
+# Jupyter checkpoints
+.ipynb_checkpoints/
+
+# VS Code settings
+.vscode/
+
+# dotenv secret file
+.env
+
+# OS files
+.DS_Store
+Thumbs.db

README.md

@@ -0,0 +1,81 @@
+# Freakquency — Disease Risk Prediction & AI Recommendations
+
+This repository contains a Streamlit web app that predicts a user's risk category for chronic diseases based on a short questionnaire and then provides AI-generated lifestyle recommendations. The app also includes a small rule-based chatbot for quick health tips.
+
+Working Figma Prototype: https://www.figma.com/proto/6WYEfSgLrciKueygaTbeCf/freakquency?type=design&node-id=1-8950&t=Hj6XFSN527rUBWYw-0&scaling=scale-down&page-id=0%3A1&starting-point-node-id=1%3A8950
+
+## What it does
+- Collects user inputs (age, gender, BMI, blood pressure, cholesterol, glucose, physical activity, smoking, alcohol intake, family history).
+- Uses a pre-trained machine learning model (`model.pkl`) and a pre-fitted scaler (`scaler.pkl`) to predict one of several disease categories (e.g., Healthy, Diabetes, Cardiovascular Disorder, Cancer, Multi-condition Cases).
+- Sends a short prompt to OpenAI to generate 3 practical lifestyle recommendations based on the predicted disease category.
+- Provides a minimal keyword-based chatbot for quick answers about exercise, diet, blood pressure, cholesterol, and glucose.
+
+## Files of interest
+- `app.py` — Streamlit application entrypoint.
+- `model.pkl` — Trained scikit-learn model used for prediction.
+- `scaler.pkl` — Feature scaler used to transform input features before prediction.
+- `chronic_disease_dataset.csv` — Dataset used during development (optional for running the app).
+- `llm_recommender.pkl` — (Optional) additional recommender artifact.
+- `requirements.txt` — Python dependencies.
+
+## Requirements
+- Python 3.8+ recommended
+- The dependencies in `requirements.txt`. Main packages include: `streamlit`, `pandas`, `numpy`, `scikit-learn`, `joblib`, `openai`.
+
+## Setup (Windows PowerShell)
+
+1. Create and activate a virtual environment (recommended):
+
+```powershell
+python -m venv .venv
+.\.venv\Scripts\Activate.ps1
+```
+
+2. Install dependencies:
+
+```powershell
+python -m pip install --upgrade pip
+pip install -r requirements.txt
+```
+
+3. Add your OpenAI API key to an environment variable. The app expects `OPENAI_API_KEY` to be available (the project uses `python-dotenv` if you prefer a `.env` file):
+
+```powershell
+$env:OPENAI_API_KEY = 'sk-...'
+# or create a .env file in the same folder as app.py with:
+# OPENAI_API_KEY=sk-...
+```
+
+Note: If you don't intend to use OpenAI (no API key), the app still works for predictions and the built-in fallback chatbot will answer some keyword queries. However, recommendations from OpenAI will not be available.
+
+## Run the Streamlit app
+
+From the `submissions/Freakquency/code/` folder run:
+
+```powershell
+streamlit run app.py
+```
+
+This will open the app in your default browser (or show a local URL in the terminal such as http://localhost:8501).
+
+## Usage
+1. Fill the questionnaire and click `Predict`.
+2. The app will display the predicted disease category and a short description.
+3. If OpenAI is configured, it will show 3 AI-generated lifestyle recommendations. Otherwise fallback recommendations are shown for certain keywords.
+4. Use the text input at the bottom to ask the small rule-based chatbot questions (e.g., "exercise", "diet").
+
+## Troubleshooting
+- If `model.pkl` or `scaler.pkl` are missing, the app will fail at startup. Make sure both files are in the same folder as `app.py`.
+- If you see OpenAI authentication errors, confirm that `OPENAI_API_KEY` is set correctly in your environment or `.env`.
+- If Streamlit fails to start, ensure your Python path and virtual environment are correct and that `streamlit` is installed in the active environment.
+
+## Security & Privacy
+- Do not commit your OpenAI API key to version control. Use environment variables or a local `.env` file excluded from source control.
+- This app is for demo/educational purposes only and does not replace professional medical advice.
+
+## License
+This project was created for the Oct-4 Hackathon 2025. Check the project or team repository for license details.
+
+---
+If you want, I can also add a brief CONTRIBUTING or LICENSE file, or update the README with screenshots and example inputs. What would you like next?
+

app.py

@@ -0,0 +1,135 @@
+import streamlit as st
+import pandas as pd
+import numpy as np
+import joblib
+import os
+from dotenv import load_dotenv
+from openai import OpenAI
+
+# --- Load environment variables from .env ---
+load_dotenv()
+
+# --- Initialize OpenAI client ---
+client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
+
+# --- Load trained model and scaler ---
+model = joblib.load("model.pkl")
+scaler = joblib.load("scaler.pkl")
+
+# --- Disease names and descriptions ---
+disease_info = {
+    0: {"name": "Healthy", "description": "No major chronic conditions detected. Maintain a balanced lifestyle to stay healthy."},
+    1: {"name": "Diabetes", "description": "Elevated glucose levels detected. Monitor diet, exercise regularly, and consult your doctor."},
+    2: {"name": "Cardiovascular Disorder", "description": "Potential risk of heart or vascular conditions. Keep track of blood pressure, cholesterol, and physical activity."},
+    3: {"name": "Cancer", "description": "Risk factors indicate possible cancer conditions. Regular screenings and medical consultation are advised."},
+    4: {"name": "Multi-condition Cases", "description": "Multiple risk factors detected. Follow a comprehensive health plan and consult a healthcare professional."}
+}
+
+# --- Minimal chatbot responses (fallback) ---
+chat_responses = {
+    "exercise": "Regular exercise helps reduce risk for multiple diseases.",
+    "diet": "A balanced diet can reduce cholesterol and glucose levels.",
+    "blood pressure": "Monitoring your blood pressure regularly is important.",
+    "cholesterol": "Check cholesterol levels regularly and maintain a healthy diet.",
+    "glucose": "Keep glucose levels in check and maintain balanced meals."
+}
+
+def generate_recommendations(disease_name, description):
+    prompt = f"""
+    The user has been predicted to have: {disease_name}.
+    Description: {description}.
+    Suggest 3 practical lifestyle recommendations for this user.
+    Keep them short, actionable, and easy to understand.
+    Format each recommendation as a bullet point.
+    """
+    response = client.chat.completions.create(
+        model="gpt-4o-mini",   # you can switch to gpt-4 or gpt-3.5 if needed
+        messages=[
+            {"role": "system", "content": "You are a helpful health assistant."},
+            {"role": "user", "content": prompt}
+        ],
+        max_tokens=200,
+        temperature=0.7
+    )
+
+    text = response.choices[0].message.content.strip()
+
+    # Clean and split into neat bullet points
+    recs = [
+        line.strip("-•1234567890. ").strip()
+        for line in text.split("\n")
+        if line.strip()
+    ]
+    return recs
+
+# --- Streamlit UI ---
+st.title("Freakquency - Disease Risk Prediction & AI Recommendations")
+st.write("Fill out the questionnaire below to see your predicted disease and AI-generated lifestyle recommendations.")
+
+# --- Streamlit Form ---
+with st.form("health_form"):
+    age = st.number_input("Age", min_value=0, max_value=120, value=30)
+    gender = st.selectbox("Gender", ["Male", "Female", "Other"])
+    bmi = st.number_input("BMI", min_value=10.0, max_value=50.0, value=25.0)
+    blood_pressure = st.number_input("Blood Pressure", min_value=50, max_value=200, value=120)
+    cholesterol_level = st.number_input("Cholesterol Level", min_value=100, max_value=400, value=180)
+    glucose_level = st.number_input("Glucose Level", min_value=50, max_value=300, value=100)
+    physical_activity = st.number_input("Physical Activity (hours/week)", min_value=0, max_value=50, value=2)
+    smoking_status = st.selectbox("Smoking Status", ["Never", "Former", "Current"])
+    alcohol_intake = st.selectbox("Alcohol Intake", ["None", "Moderate", "High"])
+    family_history = st.selectbox("Family History of Disease?", ["No", "Yes"])
+
+    submit_button = st.form_submit_button(label="Predict")
+
+# --- Preprocess and Predict ---
+if submit_button:
+    # Map categorical values to numeric
+    gender_map = {"Male": 0, "Female": 1, "Other": 2}
+    smoking_map = {"Never": 0, "Former": 1, "Current": 2}
+    alcohol_map = {"None": 0, "Moderate": 1, "High": 2}
+    family_map = {"No": 0, "Yes": 1}
+
+    input_data = np.array([[ 
+        age,
+        gender_map[gender],
+        bmi,
+        blood_pressure,
+        cholesterol_level,
+        glucose_level,
+        physical_activity,
+        smoking_map[smoking_status],
+        alcohol_map[alcohol_intake],
+        family_map[family_history]
+    ]])
+
+    # Scale features
+    input_scaled = scaler.transform(input_data)
+
+    # Predict disease class
+    prediction = model.predict(input_scaled)[0]
+    disease = disease_info[prediction]
+
+    # Display prediction in a card-like format
+    with st.container():
+        st.markdown("### Predicted Disease Category")
+        st.markdown(f"**{disease['name']}**")
+        st.write(disease["description"])
+
+        st.markdown("### AI-Generated Lifestyle Recommendations")
+        ai_recs = generate_recommendations(disease['name'], disease['description'])
+        for rec in ai_recs:
+            st.write(f"- {rec}")
+
+# --- Minimal Chatbot ---
+st.subheader("Ask a health-related question:")
+user_input = st.text_input("Type your question here...")
+
+if user_input:
+    response_given = False
+    for key, resp in chat_responses.items():
+        if key in user_input.lower():
+            st.write(resp)
+            response_given = True
+            break
+    if not response_given:
+        st.write("Sorry, I don't have advice for that yet. Try asking about exercise, diet, or blood pressure.")