#!/usr/bin/env python
# coding: utf-8

# In[15]:


import csv
import pandas as pd
import numpy as np
from sklearn.cluster import KMeans
import imdb
import random
import gradio as gr


# In[2]:


file1=open("movies.csv",'r')
file2=open("movies3.csv",'r')
csvreader1 = csv.reader(file1)
csvreader2 = csv.reader(file2)


# In[3]:


pd_dict={'movie':[],'time_minute':[],'imdb_rating':[],'Action':[],'Adventure':[],'Fantasy':[],
         'Sci-Fi':[],'Animation':[],'Comedy':[],'Family':[],'Mystery':[],'Romance':[],'Drama':[],
         'Crime':[],'Thriller':[],'War':[],'Musical':[],'Biography':[]}

lst=['Action','Adventure','Fantasy','Sci-Fi','Animation','Comedy',
     'Family','Mystery','Romance','Drama','Crime','Thriller','War','Musical','Biography']


# In[4]:


j=0
for row in csvreader2:
    if j==0:
        j=1
        continue
    for i in range(len(lst)):
        pd_dict[lst[i]].append(int(row[i+1]))
j=0
for row in csvreader1:
    if j==0:
        j=1
        continue
    if j==250:
        break
    pd_dict['movie'].append(row[1])
    pd_dict['time_minute'].append(int(row[3].split()[0]))
    pd_dict['imdb_rating'].append(float(row[4]))
    j+=1


# In[5]:


df=pd.DataFrame(pd_dict)
X=np.array(df.drop(['movie'], axis = 1))


# In[6]:


kmeans = KMeans(max_iter=300).fit(X)


# In[9]:


movie_name="Oppenheimer"
ia=imdb.IMDb()
items = ia.search_movie(movie_name)
code = items[0].getID()
print(code)
series = ia.get_movie(code)
genre = series.data['genres']
time_minute=series.data['runtimes']
rating=series.data['rating']
x=[int(time_minute[0]),float(rating)]
for i in lst:
    if i in genre:
        x.append(1)
    else:
        x.append(0)
print(x)


# In[13]:


i=random.randint(1,249)
test_x=X[i].reshape(1,-1)
x=np.array(x)
x=x.reshape(1,-1)
# print(kmeans.predict(test_x))
# print(kmeans.predict(x))
pred=kmeans.predict(x)
pred_test=kmeans.predict(test_x)
while pred!=pred_test:
    i=random.randint(1,249)
    test_x=X[i].reshape(1,-1)
    x=np.array(x)
    x=x.reshape(1,-1)
    pred_test=kmeans.predict(test_x)
print('Another movie like',movie_name,'is:',df['movie'][i])


# In[16]:


def movie(txt):
    ia=imdb.IMDb()
    items = ia.search_movie(movie_name)
    code = items[0].getID()
    series = ia.get_movie(code)
    genre = series.data['genres']
    time_minute=series.data['runtimes']
    rating=series.data['rating']
    x=[int(time_minute[0]),float(rating)]
    for i in lst:
        if i in genre:
            x.append(1)
        else:
            x.append(0)
            
    i=random.randint(1,249)
    test_x=X[i].reshape(1,-1)
    x=np.array(x)
    x=x.reshape(1,-1)
    # print(kmeans.predict(test_x))
    # print(kmeans.predict(x))
    pred=kmeans.predict(x)
    pred_test=kmeans.predict(test_x)
    while pred!=pred_test:
        i=random.randint(1,249)
        test_x=X[i].reshape(1,-1)
        x=np.array(x)
        x=x.reshape(1,-1)
        pred_test=kmeans.predict(test_x)
    return df['movie'][i]

iface = gr.Interface(fn=movie, inputs="text", outputs="text")
iface.launch()


# In[ ]: