Shrikrishna/Movie_Recommender_system

1import streamlit as st2import pickle3import requests4import pandas as pd5 6footer="""<style>7a:link , a:visited{8color: black;9background-color: transparent;10}11 12a:hover,  a:active {13color: red;14background-color: transparent;15}16 17.footer {18position: fixed;19left: 0;20bottom: 0;21width: 100%;22background-color: white;23color: black;24text-align: center;25}26</style>27<div class="footer">28<p>Developed with <span style ='color:red'>❤</span> by <a href="https://shrikrishnaparab.tech/" target="_blank">Shrikrishna Parab</a></p>29</div>30"""31 32def fetch_poster(movie_id):33    url = "https://api.themoviedb.org/3/movie/{}?api_key=8265bd1679663a7ea12ac168da84d2e8&language=en-US".format(movie_id)34    data = requests.get(url)35    data = data.json()36    poster_path = data['poster_path']37    full_path = "https://image.tmdb.org/t/p/w500/" + poster_path38    return full_path39 40def get_popular(qualified):41    top_5 = qualified.head(5)42    return top_543 44 45def top_genre_based_movies(genre, percentile=0.95):46    df = genre_df[genre_df['genres'].str.contains(genre)]47    vote_counts = df['vote_count'].astype('int')48    vote_averages = df['vote_average'].astype('int')49    C = vote_averages.mean()50    m = vote_counts.quantile(percentile)51    qualified = df[(df['vote_count'] >= m)][['movie_id', 'title', 'vote_count', 'vote_average', 'genres']]52    qualified['vote_count'] = qualified['vote_count'].astype('int')53    qualified['vote_average'] = qualified['vote_average'].astype('int')54    qualified['wr'] = qualified.apply(55        lambda x: (x['vote_count'] / (x['vote_count'] + m) * x['vote_average']) + (m / (m + x['vote_count']) * C),56        axis=1)57    qualified = qualified.sort_values('wr', ascending=False).head(250)58    return qualified59 60def recommend(movie):61    index = movies[movies['title'] == movie].index[0]62    distances = sorted(list(enumerate(similarity[index])), reverse=True, key=lambda x: x[1])63    recommended_movie_names = []64    recommended_movie_posters = []65    for i in distances[1:6]:66        # fetch the movie poster67        movie_id = movies.iloc[i[0]].movie_id68        recommended_movie_posters.append(fetch_poster(movie_id))69        recommended_movie_names.append(movies.iloc[i[0]].title)70 71    return recommended_movie_names,recommended_movie_posters72 73 74st.title("Movie Recommender System")75 76movies = pickle.load(open('movie_list.pkl','rb'))77similarity = pickle.load(open('similarity.pkl','rb'))78all_movies = pickle.load(open('movies_df.pkl','rb'))79top_popular = pickle.load(open('top_popular.pkl','rb'))80 81s = all_movies.apply(lambda x: pd.Series(x['genres']),axis=1).stack().reset_index(level=1, drop=True)82s.name = 'genres'83genre_df = all_movies.drop('genres', axis=1).join(s)84 85movie_list = movies['title'].values86option_selected = st.selectbox(87    'Type or Select Movie Name from Dropdown',88    movie_list89)90 91genre_list = ['Action','Romance','Adventure','Science Fiction','Comedy']92genre_selected = st.selectbox(93    'Type or Select Genre from Dropdown',94    genre_list95)96 97if st.button('Show Recommendation'):98    recommended_movie_names, recommended_movie_posters = recommend(option_selected)99    top_popular_movies = get_popular(top_popular)100    st.header("Movies Based on Content: Similar Movies")101    col1, col2, col3, col4, col5 = st.columns(5)102    with col1:103        st.image(recommended_movie_posters[0], caption=recommended_movie_names[0])104    with col2:105        st.image(recommended_movie_posters[1], caption=recommended_movie_names[1])106 107    with col3:108        st.image(recommended_movie_posters[2], caption=recommended_movie_names[2])109    with col4:110        st.image(recommended_movie_posters[3], caption=recommended_movie_names[3])111    with col5:112        st.image(recommended_movie_posters[4], caption=recommended_movie_names[4])113 114    st.header("Movies Based on Popularity: Top Popular")115    popular = []116    for row in top_popular_movies.loc[:,['title','movie_id']].values:117        popular.append(row)118    col6, col7, col8, col9, col10 = st.columns(5)119    with col6:120        full_path = fetch_poster(popular[0][1])121        st.image(full_path, caption=popular[0][0])122    with col7:123        full_path = fetch_poster(popular[1][1])124        st.image(full_path, caption=popular[1][0])125    with col8:126        full_path = fetch_poster(popular[2][1])127        st.image(full_path, caption=popular[2][0])128    with col9:129        full_path = fetch_poster(popular[3][1])130        st.image(full_path, caption=popular[3][0])131    with col10:132        full_path = fetch_poster(popular[4][1])133        st.image(full_path, caption=popular[4][0])134 135 136    st.header("Movies Based on Genre: Top "+str(genre_selected)+" Movies")137    top_gener_based = top_genre_based_movies(genre_selected).head(5)138    genre_popular = []139    for row in top_gener_based.loc[:, ['title', 'movie_id']].values:140        genre_popular.append(row)141    col11, col12, col13, col14, col15 = st.columns(5)142    with col11:143        full_path = fetch_poster(genre_popular[0][1])144        st.image(full_path, caption=genre_popular[0][0])145    with col12:146        full_path = fetch_poster(genre_popular[1][1])147        st.image(full_path, caption=genre_popular[1][0])148    with col13:149        full_path = fetch_poster(genre_popular[2][1])150        st.image(full_path, caption=genre_popular[2][0])151    with col14:152        full_path = fetch_poster(genre_popular[3][1])153        st.image(full_path, caption=genre_popular[3][0])154    with col15:155        full_path = fetch_poster(genre_popular[4][1])156        st.image(full_path, caption=genre_popular[4][0])157 158    159 160 161 162 163st.markdown(footer,unsafe_allow_html=True)164