recipes_recommendation.py

import streamlit as st
import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity

# For AI model (replace with your chosen model)
import tensorflow as tf

# Replace with your recipe data
recipes = pd.read_csv("recipes.csv")

def recommend_recipes(user_preferences, available_ingredients):
    # Preprocess user inputs and recipe data
    # ...

    # Create TF-IDF vectors
    tfidf_vectorizer = TfidfVectorizer(stop_words='english')
    tfidf_matrix = tfidf_vectorizer.fit_transform(recipes['ingredients'])

    # Calculate cosine similarity
    cosine_similarities = cosine_similarity(tfidf_matrix)

    # Find the most similar recipes based on user inputs
    # ...

    return recommended_recipes

def main():
    st.title("AI Recipe Recommender")

    # User inputs
    dietary_restrictions = st.multiselect("Dietary Restrictions:", ["Vegetarian", "Vegan", "Gluten-free", "Dairy-free"])
    cuisine = st.selectbox("Cuisine:", ["Any", "Italian", "Mexican", "Indian", "Chinese"])
    ingredients = st.text_input("Available Ingredients (comma-separated):")
    allergies = st.text_input("Allergies (comma-separated):")

    if st.button("Recommend Recipes"):
        user_preferences = {
            "dietary_restrictions": dietary_restrictions,
            "cuisine": cuisine,
            "ingredients": ingredients.split(","),
            "allergies": allergies.split(",")
        }

        recommended_recipes = recommend_recipes(user_preferences, ingredients.split(","))

        # Display recommended recipes with details and potential substitutions
        for recipe in recommended_recipes:
            st.subheader(recipe['title'])
            st.write(recipe['description'])
            # ... display ingredients, instructions, etc.
            st.write("Potential Substitutions:")
            # ... display potential substitutions based on user preferences and allergies

if __name__ == "__main__":
    main()