app.py

import gradio as gr
import numpy as np

# Data: Nutritional Information and Cost by City for all dishes
dish_data = {
    "Idli with Vegetable Sambar": {
        "Energy (kcal)": 100,
        "Protein (g)": 5.5,
        "Fat (g)": 1.5,
        "Carbohydrate (g)": 17.0,
        "Fiber (g)": 3.0,
        "Calcium (mg)": 35,
        "Iron (mg)": 1.0,
        "Vitamin C (mg)": 8.0,
        "Chennai": {"Street": "20-60", "Local": "40-90", "Upscale": "80-180"},
        "Bengaluru": {"Street": "25-60", "Local": "50-100", "Upscale": "80-200"},
        "Hyderabad": {"Street": "20-50", "Local": "40-90", "Upscale": "80-180"},
        "New Delhi": {"Street": "30-70", "Local": "60-120", "Upscale": "90-200"}
    },
    "Khichdi": {
        "Energy (kcal)": 120,
        "Protein (g)": 4.0,
        "Fat (g)": 2.5,
        "Carbohydrate (g)": 20.0,
        "Fiber (g)": 2.5,
        "Calcium (mg)": 40,
        "Iron (mg)": 1.0,
        "Vitamin C (mg)": 5.0,
        "Chennai": {"Street": "30-60", "Local": "50-120", "Upscale": "120-200"},
        "Bengaluru": {"Street": "35-70", "Local": "60-130", "Upscale": "120-220"},
        "Hyderabad": {"Street": "30-60", "Local": "50-110", "Upscale": "110-200"},
        "New Delhi": {"Street": "40-80", "Local": "70-140", "Upscale": "130-250"}
    },
    "Tandoori Chicken": {
        "Energy (kcal)": 150,
        "Protein (g)": 18.0,
        "Fat (g)": 7.0,
        "Carbohydrate (g)": 3.0,
        "Fiber (g)": 0.5,
        "Calcium (mg)": 15,
        "Iron (mg)": 1.2,
        "Vitamin C (mg)": 1.5,
        "Chennai": {"Street": "100-150", "Local": "150-300", "Upscale": "250-500"},
        "Bengaluru": {"Street": "110-160", "Local": "160-320", "Upscale": "270-550"},
        "Hyderabad": {"Street": "100-150", "Local": "150-300", "Upscale": "250-500"},
        "New Delhi": {"Street": "120-180", "Local": "180-350", "Upscale": "300-600"}
    },
    "Palak Paneer": {
        "Energy (kcal)": 140,
        "Protein (g)": 7.5,
        "Fat (g)": 10.0,
        "Carbohydrate (g)": 6.0,
        "Fiber (g)": 3.0,
        "Calcium (mg)": 200,
        "Iron (mg)": 3.0,
        "Vitamin C (mg)": 15.0,
        "Chennai": {"Street": "60-100", "Local": "100-200", "Upscale": "180-350"},
        "Bengaluru": {"Street": "70-110", "Local": "110-220", "Upscale": "200-380"},
        "Hyderabad": {"Street": "60-100", "Local": "100-200", "Upscale": "180-350"},
        "New Delhi": {"Street": "80-120", "Local": "120-250", "Upscale": "200-400"}
    },
    "Raita": {
        "Energy (kcal)": 60,
        "Protein (g)": 3.5,
        "Fat (g)": 2.0,
        "Carbohydrate (g)": 6.5,
        "Fiber (g)": 0.5,
        "Calcium (mg)": 100,
        "Iron (mg)": 0.5,
        "Vitamin C (mg)": 2.0,
        "Chennai": {"Street": "20-40", "Local": "50-80", "Upscale": "100-150"},
        "Bengaluru": {"Street": "25-50", "Local": "55-85", "Upscale": "110-160"},
        "Hyderabad": {"Street": "20-40", "Local": "50-80", "Upscale": "100-150"},
        "New Delhi": {"Street": "30-50", "Local": "60-100", "Upscale": "120-180"}
    },
    "Rajma": {
        "Energy (kcal)": 140,
        "Protein (g)": 7.5,
        "Fat (g)": 5.0,
        "Carbohydrate (g)": 20.0,
        "Fiber (g)": 6.0,
        "Calcium (mg)": 50,
        "Iron (mg)": 3.5,
        "Vitamin C (mg)": 4.0,
        "Chennai": {"Street": "50-80", "Local": "100-150", "Upscale": "180-300"},
        "Bengaluru": {"Street": "60-90", "Local": "110-160", "Upscale": "200-320"},
        "Hyderabad": {"Street": "50-80", "Local": "100-150", "Upscale": "180-300"},
        "New Delhi": {"Street": "70-100", "Local": "120-180", "Upscale": "220-350"}
    },
    "Baingan Bharta": {
        "Energy (kcal)": 70,
        "Protein (g)": 2.5,
        "Fat (g)": 3.0,
        "Carbohydrate (g)": 10.0,
        "Fiber (g)": 4.0,
        "Calcium (mg)": 30,
        "Iron (mg)": 0.7,
        "Vitamin C (mg)": 6.0,
        "Chennai": {"Street": "40-70", "Local": "80-120", "Upscale": "150-250"},
        "Bengaluru": {"Street": "45-80", "Local": "90-130", "Upscale": "160-270"},
        "Hyderabad": {"Street": "40-70", "Local": "80-120", "Upscale": "150-250"},
        "New Delhi": {"Street": "50-90", "Local": "100-150", "Upscale": "180-280"}
    },
    "Besan Chilla": {
        "Energy (kcal)": 180,
        "Protein (g)": 8.0,
        "Fat (g)": 7.0,
        "Carbohydrate (g)": 20.0,
        "Fiber (g)": 5.0,
        "Calcium (mg)": 30,
        "Iron (mg)": 2.5,
        "Vitamin C (mg)": 1.0,
        "Chennai": {"Street": "20-40", "Local": "50-80", "Upscale": "100-150"},
        "Bengaluru": {"Street": "25-45", "Local": "55-85", "Upscale": "110-160"},
        "Hyderabad": {"Street": "20-40", "Local": "50-80", "Upscale": "100-150"},
        "New Delhi": {"Street": "30-50", "Local": "60-90", "Upscale": "120-180"}
    },
    "Masoor Dal": {
        "Energy (kcal)": 110,
        "Protein (g)": 7.5,
        "Fat (g)": 2.0,
        "Carbohydrate (g)": 16.0,
        "Fiber (g)": 5.5,
        "Calcium (mg)": 25,
        "Iron (mg)": 3.0,
        "Vitamin C (mg)": 4.0,
        "Chennai": {"Street": "40-70", "Local": "80-120", "Upscale": "150-250"},
        "Bengaluru": {"Street": "45-75", "Local": "85-130", "Upscale": "160-270"},
        "Hyderabad": {"Street": "40-70", "Local": "80-120", "Upscale": "150-250"},
        "New Delhi": {"Street": "50-90", "Local": "100-150", "Upscale": "180-280"}
    },
    "Upma": {
        "Energy (kcal)": 150,
        "Protein (g)": 4.0,
        "Fat (g)": 6.0,
        "Carbohydrate (g)": 25.0,
        "Fiber (g)": 3.0,
        "Calcium (mg)": 30,
        "Iron (mg)": 1.0,
        "Vitamin C (mg)": 3.0,
        "Chennai": {"Street": "20-50", "Local": "50-100", "Upscale": "100-200"},
        "Bengaluru": {"Street": "25-60", "Local": "55-110", "Upscale": "120-220"},
        "Hyderabad": {"Street": "20-50", "Local": "50-100", "Upscale": "100-200"},
        "New Delhi": {"Street": "30-60", "Local": "60-120", "Upscale": "120-250"}
    }
}

def extract_average_cost(cost_range):
    """Extracts the average cost from a cost range string."""
    costs = [int(cost) for cost in cost_range.split('-')]
    return np.mean(costs)

def normalize(values):
    """Normalizes a list of values to a scale from 0 to 1."""
    min_val = min(values)
    max_val = max(values)
    return [(v - min_val) / (max_val - min_val) for v in values]

def optimize_nutrition_and_cost(city, segment, nutrition_weight, cost_weight):
    """Optimizes dishes based on both nutrition and cost with custom weights."""
    nutrition_scores = []
    costs = []
    dishes = []

    for dish, info in dish_data.items():
        nutrition_score = (
            info["Protein (g)"] * 2 +  # Higher protein is better
            info["Fiber (g)"] * 1.5 +  # Higher fiber is better
            -info["Energy (kcal)"]  # Lower calories are better
        )
        nutrition_scores.append(nutrition_score)
        avg_cost = extract_average_cost(info[city][segment])
        costs.append(avg_cost)
        dishes.append(dish)

    # Normalize both nutrition and cost
    normalized_nutrition_scores = normalize(nutrition_scores)
    normalized_costs = normalize(costs)
    
    # Calculate combined score (higher is better)
    combined_scores = [
        (nutrition_weight * n_score + cost_weight * (1 - c_score))
        for n_score, c_score in zip(normalized_nutrition_scores, normalized_costs)
    ]

    # Sort by combined score
    ranked_dishes = sorted(zip(dishes, combined_scores), key=lambda x: x[1], reverse=True)

    return ranked_dishes

def display_optimized_ranking(city, segment, nutrition_weight, cost_weight):
    """Displays the optimized ranking of dishes based on nutrition and cost."""
    ranked_dishes = optimize_nutrition_and_cost(city, segment, nutrition_weight, cost_weight)
    output = ""
    
    for i, (dish, combined_score) in enumerate(ranked_dishes, 1):
        output += f"{i}. {dish} - Combined Score: {combined_score:.2f}\n"
    
    return output

# Gradio Interface
def create_interface():
    cities = ["Chennai", "Bengaluru", "Hyderabad", "New Delhi"]
    segments = ["Street", "Local", "Upscale"]
    
    with gr.Blocks() as demo:
        gr.Markdown("# Healthy Indian Dishes: Optimized by Cost and Nutrition")
        with gr.Row():
            city_selector = gr.Dropdown(choices=cities, label="Select a City", value="Chennai")
            segment_selector = gr.Dropdown(choices=segments, label="Select an Eateries Segment", value="Street")
        with gr.Row():
            nutrition_weight_slider = gr.Slider(0, 1, value=0.7, label="Nutrition Weight")
            cost_weight_slider = gr.Slider(0, 1, value=0.3, label="Cost Weight")
        with gr.Row():
            optimized_ranking_output = gr.Textbox(label="Optimized Dish Ranking", interactive=False)
        
        # Update the output based on the selections and weight sliders
        def update_ranking(city, segment, nutrition_weight, cost_weight):
            return display_optimized_ranking(city, segment, nutrition_weight, cost_weight)
        
        inputs = [city_selector, segment_selector, nutrition_weight_slider, cost_weight_slider]
        city_selector.change(fn=update_ranking, inputs=inputs, outputs=optimized_ranking_output)
        segment_selector.change(fn=update_ranking, inputs=inputs, outputs=optimized_ranking_output)
        nutrition_weight_slider.change(fn=update_ranking, inputs=inputs, outputs=optimized_ranking_output)
        cost_weight_slider.change(fn=update_ranking, inputs=inputs, outputs=optimized_ranking_output)
        
        gr.Row(city_selector, segment_selector, nutrition_weight_slider, cost_weight_slider, optimized_ranking_output)
    
    return demo

# Launch the interface
demo = create_interface()
demo.launch(share=True)