app.py

import pickle
import streamlit as st
import requests
import pandas as pd

# set page setting
st.set_page_config(page_title='TopMovies')

# set history var
if 'history' not in st.session_state:
    st.session_state.history = []

# import preprocessed data
data = pd.read_csv("./data/tags.csv")

# import similarity (to be cached)
def importSim(filename):
    sim = pickle.load(open(filename, 'rb'))
    return sim

similarity = importSim('similarity.pkl')

# recommender function
def recommend_image(movie, sim):
    poster = []
    plot = []
    # index from dataframe
    index = data[data['title'] == movie].index[0]
    dist = dict(enumerate(sim[index]))
    dist = dict(sorted(dist.items(), reverse=True, key = lambda item: item[1]))
    #index from 1 because the forst is the movie itself
    cnt = 0
    for key in dist: 
        cnt = cnt+1
        if cnt < 11:
            title = data.iloc[key].title
            posterRes, plotRes = get_poster_plot(title) 
            poster.append(posterRes)
            plot.append(plotRes)
        else:
            break
        
    return poster[1:], plot[1:]

# get poster
def get_poster_plot(title):
    r = requests.get("http://www.omdbapi.com/?i=tt3896198&apikey=37765f04&t=" + title).json()
    posterElement = r["Poster"]
    plotElement = r["Plot"]
    return posterElement, plotElement

# update last viewed list
def update_las_viewed():
    if len(st.session_state.history) > 3:
        st.session_state.history.pop()

# sidebar
st.sidebar.write("""
This is a content based recommender system. Pick a movie from the list or search for it and then wait for the reccomendations.
You will get six movies, posters and plots.
""")

# title
st.write("# Movie Recommendation System")
st.write("Pick a movie from the list and enjoy some new stuffs!")

# select box
title = st.selectbox("", data["title"])
if title not in st.session_state.history:
    st.session_state.history.insert(0, title)
update_las_viewed()

# recommend
with st.spinner("Getting the best movies..."):
    recs, plots = recommend_image(title, similarity)

# recommendation cols
st.write("## What to watch next....")
col1, col2, col3 = st.columns(3)
with col1:
    st.image(recs[0])
    st.write(plots[0])
with col2:
    st.image(recs[1])
    st.write(plots[1])
with col3:
    st.image(recs[2])
    st.write(plots[2])

col4, col5, col6 = st.columns(3)
with col4:
    st.image(recs[3])
    st.write(plots[3])
with col5:
    st.image(recs[4])
    st.write(plots[4])
with col6:
    st.image(recs[5])
    st.write(plots[5])

col7, col8, col9 = st.columns(3)
with col7:
    st.image(recs[6])
    st.write(plots[6])
with col8:
    st.image(recs[7])
    st.write(plots[7])
with col9:
    st.image(recs[8])
    st.write(plots[8])

# last viewed
st.write("## Last viewed:")
r1, r2, r3 = st.columns(3)
with r1:
    try:
        st.image(get_poster_plot(st.session_state.history[0])[0])
    except IndexError:
        pass
    
with r2:
    try:
        st.image(get_poster_plot(st.session_state.history[1])[0])
    except IndexError:
        pass
    
with r3:
    try:
        st.image(get_poster_plot(st.session_state.history[2])[0])
    except IndexError:
        pass