In [1]:
import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt 
import seaborn as sns
import ast
import matplotlib.image as mpimg
%matplotlib inline
from wordcloud import WordCloud, STOPWORDS
from IPython.core.interactiveshell import InteractiveShell
InteractiveShell.ast_node_interactivity = "all"
from IPython import display
import matplotlib.image as mpimg

Kino industrijos analizė¶

In [3]:
from pathlib import Path
gifPath = Path("D:\\giphy (3).gif")
with open(gifPath,'rb') as f:
    display.Image(data=f.read(), format='png')
Out[3]:

Duomenų rinkinio apžvalga¶

Duomenys yra paimti iš https://www.kaggle.com/ - didžiausios pasaulyje duomenų mokslo bendruomenės. Duomenų rinkinį sudaro filmai, išleisti iki 2017 m. liepos mėn, suirinkti iš „MovieLens“ internetinės rekomendacijų sistemos, kuri rekomenduoja filmus žiūrėti savo vartotojams, atsižvelgiant į jų pasirinktus filmus, filtruojant narių filmų įvertinimus ir filmų peržiūras.

Pagrindiniai iškelti klausimai¶

  1. Kurie filmai daugiausiai uždirbo? Didžiausio biudžeto filmai?
  2. Kurie filmai buvo nuostolingi?
  3. Kurie režisieriai yra sukūrę daugiausiai filmų?
  4. Kokio žanro filmų yra kuriama daugiausiai?
  5. Kokie žodžiai yra dažniausiai pasikartojantys filmų pavadinimuose?
  6. Kurie filmai TMDB bendruomenės įvertinti geriausiai?
  7. Kaip kito filmų ilgis bėgant metams?

Hipotezė- Kino pramonę valdo vyrai?

Duomenų valymas ir tvarkymas¶

In [4]:
meta = pd.read_csv('D:\\movies_metadata.csv')
meta.head()

D:\anaconda3\lib\site-packages\IPython\core\interactiveshell.py:3444: DtypeWarning: Columns (10) have mixed types.Specify dtype option on import or set low_memory=False.
  exec(code_obj, self.user_global_ns, self.user_ns)
Out[4]:
adult belongs_to_collection budget genres homepage id imdb_id original_language original_title overview ... release_date revenue runtime spoken_languages status tagline title video vote_average vote_count
0 False {'id': 10194, 'name': 'Toy Story Collection', ... 30000000 [{'id': 16, 'name': 'Animation'}, {'id': 35, '... http://toystory.disney.com/toy-story 862 tt0114709 en Toy Story Led by Woody, Andy's toys live happily in his ... ... 1995-10-30 373554033.0 81.0 [{'iso_639_1': 'en', 'name': 'English'}] Released NaN Toy Story False 7.7 5415.0
1 False NaN 65000000 [{'id': 12, 'name': 'Adventure'}, {'id': 14, '... NaN 8844 tt0113497 en Jumanji When siblings Judy and Peter discover an encha... ... 1995-12-15 262797249.0 104.0 [{'iso_639_1': 'en', 'name': 'English'}, {'iso... Released Roll the dice and unleash the excitement! Jumanji False 6.9 2413.0
2 False {'id': 119050, 'name': 'Grumpy Old Men Collect... 0 [{'id': 10749, 'name': 'Romance'}, {'id': 35, ... NaN 15602 tt0113228 en Grumpier Old Men A family wedding reignites the ancient feud be... ... 1995-12-22 0.0 101.0 [{'iso_639_1': 'en', 'name': 'English'}] Released Still Yelling. Still Fighting. Still Ready for... Grumpier Old Men False 6.5 92.0
3 False NaN 16000000 [{'id': 35, 'name': 'Comedy'}, {'id': 18, 'nam... NaN 31357 tt0114885 en Waiting to Exhale Cheated on, mistreated and stepped on, the wom... ... 1995-12-22 81452156.0 127.0 [{'iso_639_1': 'en', 'name': 'English'}] Released Friends are the people who let you be yourself... Waiting to Exhale False 6.1 34.0
4 False {'id': 96871, 'name': 'Father of the Bride Col... 0 [{'id': 35, 'name': 'Comedy'}] NaN 11862 tt0113041 en Father of the Bride Part II Just when George Banks has recovered from his ... ... 1995-02-10 76578911.0 106.0 [{'iso_639_1': 'en', 'name': 'English'}] Released Just When His World Is Back To Normal... He's ... Father of the Bride Part II False 5.7 173.0

5 rows × 24 columns

In [5]:
crew = pd.read_csv('D:\\credits.csv')
crew.head()
Out[5]:
cast crew id
0 [{'cast_id': 14, 'character': 'Woody (voice)',... [{'credit_id': '52fe4284c3a36847f8024f49', 'de... 862
1 [{'cast_id': 1, 'character': 'Alan Parrish', '... [{'credit_id': '52fe44bfc3a36847f80a7cd1', 'de... 8844
2 [{'cast_id': 2, 'character': 'Max Goldman', 'c... [{'credit_id': '52fe466a9251416c75077a89', 'de... 15602
3 [{'cast_id': 1, 'character': "Savannah 'Vannah... [{'credit_id': '52fe44779251416c91011acb', 'de... 31357
4 [{'cast_id': 1, 'character': 'George Banks', '... [{'credit_id': '52fe44959251416c75039ed7', 'de... 11862
In [6]:
meta.id.drop_duplicates(keep = 'first', inplace=True)
crew.id.drop_duplicates(keep = 'first', inplace=True)
In [7]:
meta = meta.drop(meta[meta.original_language.str.count('\d') != 0].index)
crew = crew.drop(meta[meta.original_language.str.count('\d') != 0].index)
In [8]:
meta.id = meta.id.astype(int)
In [9]:
movies = meta.merge(crew, on='id')
movies.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 45527 entries, 0 to 45526
Data columns (total 26 columns):
 #   Column                 Non-Null Count  Dtype  
---  ------                 --------------  -----  
 0   adult                  45527 non-null  object 
 1   belongs_to_collection  4500 non-null   object 
 2   budget                 45527 non-null  object 
 3   genres                 45527 non-null  object 
 4   homepage               7792 non-null   object 
 5   id                     45527 non-null  int32  
 6   imdb_id                45510 non-null  object 
 7   original_language      45527 non-null  object 
 8   original_title         45527 non-null  object 
 9   overview               44574 non-null  object 
 10  popularity             45524 non-null  object 
 11  poster_path            45141 non-null  object 
 12  production_companies   45524 non-null  object 
 13  production_countries   45524 non-null  object 
 14  release_date           45440 non-null  object 
 15  revenue                45524 non-null  float64
 16  runtime                45267 non-null  float64
 17  spoken_languages       45524 non-null  object 
 18  status                 45443 non-null  object 
 19  tagline                20439 non-null  object 
 20  title                  45524 non-null  object 
 21  video                  45524 non-null  object 
 22  vote_average           45524 non-null  float64
 23  vote_count             45524 non-null  float64
 24  cast                   45527 non-null  object 
 25  crew                   45527 non-null  object 
dtypes: float64(4), int32(1), object(21)
memory usage: 9.2+ MB
In [10]:
movies_clean = movies.copy() # išsaugome sujungtą lentelę
In [11]:
movies_clean = movies_clean.drop(['imdb_id', 'tagline', 'video', 'homepage', 'adult','poster_path','belongs_to_collection','overview'], axis=1)
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 45527 entries, 0 to 45526
Data columns (total 18 columns):
 #   Column                Non-Null Count  Dtype  
---  ------                --------------  -----  
 0   budget                45527 non-null  object 
 1   genres                45527 non-null  object 
 2   id                    45527 non-null  int32  
 3   original_language     45527 non-null  object 
 4   original_title        45527 non-null  object 
 5   popularity            45524 non-null  object 
 6   production_companies  45524 non-null  object 
 7   production_countries  45524 non-null  object 
 8   release_date          45440 non-null  object 
 9   revenue               45524 non-null  float64
 10  runtime               45267 non-null  float64
 11  spoken_languages      45524 non-null  object 
 12  status                45443 non-null  object 
 13  title                 45524 non-null  object 
 14  vote_average          45524 non-null  float64
 15  vote_count            45524 non-null  float64
 16  cast                  45527 non-null  object 
 17  crew                  45527 non-null  object 
dtypes: float64(4), int32(1), object(13)
memory usage: 6.4+ MB
In [12]:
movies_clean.drop(movies_clean[movies_clean['status'] != 'Released'].index, inplace=True) # Panaikinam filmus, kurių statusas yra kitoks nei 'released'
In [13]:
movies_clean = movies_clean.drop(['status'], axis=1) # Kadangi visi filmai yra su statusu 'released' panaikinam visai šį stulpelį
In [14]:
zero_runtime = movies_clean[movies_clean['runtime'] == 0]  # Panaikinam visus filmus kuriu runtime ir vote count yra 0
movies_clean.drop(zero_runtime.index, inplace=True)

zero_vote = movies_clean[movies_clean['vote_count'] == 0]
movies_clean.drop(zero_vote.index, inplace=True)
In [15]:
movies_clean[['popularity']]=movies_clean[['popularity']].astype(str).astype(float)  # Pakeičiam duomenų tipus
movies_clean[['budget']]=movies_clean[['budget']].astype(str).astype(float)
movies_clean['release_date']=pd.to_datetime(movies_clean['release_date'])

movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 41065 entries, 0 to 45524
Data columns (total 17 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                41065 non-null  float64       
 1   genres                41065 non-null  object        
 2   id                    41065 non-null  int32         
 3   original_language     41065 non-null  object        
 4   original_title        41065 non-null  object        
 5   popularity            41065 non-null  float64       
 6   production_companies  41065 non-null  object        
 7   production_countries  41065 non-null  object        
 8   release_date          41041 non-null  datetime64[ns]
 9   revenue               41065 non-null  float64       
 10  runtime               40911 non-null  float64       
 11  spoken_languages      41065 non-null  object        
 12  title                 41065 non-null  object        
 13  vote_average          41065 non-null  float64       
 14  vote_count            41065 non-null  float64       
 15  cast                  41065 non-null  object        
 16  crew                  41065 non-null  object        
dtypes: datetime64[ns](1), float64(6), int32(1), object(9)
memory usage: 5.5+ MB
In [16]:
movies_clean[['title', 'genres', 'production_companies', 'production_countries', 'spoken_languages']].head(3)
Out[16]:
title genres production_companies production_countries spoken_languages
0 Toy Story [{'id': 16, 'name': 'Animation'}, {'id': 35, '... [{'name': 'Pixar Animation Studios', 'id': 3}] [{'iso_3166_1': 'US', 'name': 'United States o... [{'iso_639_1': 'en', 'name': 'English'}]
1 Jumanji [{'id': 12, 'name': 'Adventure'}, {'id': 14, '... [{'name': 'TriStar Pictures', 'id': 559}, {'na... [{'iso_3166_1': 'US', 'name': 'United States o... [{'iso_639_1': 'en', 'name': 'English'}, {'iso...
2 Grumpier Old Men [{'id': 10749, 'name': 'Romance'}, {'id': 35, ... [{'name': 'Warner Bros.', 'id': 6194}, {'name'... [{'iso_3166_1': 'US', 'name': 'United States o... [{'iso_639_1': 'en', 'name': 'English'}]
In [17]:
# Pasiimam iš stulpelių production_countries,production_companies,spoken_languages ir genres  tik tą informaciją kuri mums bus reikalinga
In [18]:
movies_clean['production_countries']=movies_clean.loc[movies_clean['production_countries'].notna(), 'production_countries'].str.split("'name': ").str[1].str.split("'").str[1]
In [19]:
movies_clean['production_companies']=movies_clean.loc[movies_clean['production_companies'].notna(), 'production_companies'].str.split("'name': ").str[1].str.split("'").str[1]
In [20]:
movies_clean['spoken_languages']=movies_clean.loc[movies_clean['spoken_languages'].notna(), 'spoken_languages'].str.split("'name': ").str[1].str.split("'").str[1]
In [21]:
movies_clean['genres']=movies_clean.loc[movies_clean['genres'].notna(), 'genres'].str.split("'name': ").str[1].str.split("'").str[1]
In [22]:
movies_clean['year'] = movies_clean['release_date'].dt.year
movies_clean['year'].value_counts()
Out[22]:
2014.0    1780
2013.0    1759
2015.0    1699
2012.0    1576
2011.0    1525
          ... 
1878.0       1
1874.0       1
1887.0       1
1883.0       1
1893.0       1
Name: year, Length: 133, dtype: int64
In [23]:
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 41065 entries, 0 to 45524
Data columns (total 18 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                41065 non-null  float64       
 1   genres                39663 non-null  object        
 2   id                    41065 non-null  int32         
 3   original_language     41065 non-null  object        
 4   original_title        41065 non-null  object        
 5   popularity            41065 non-null  float64       
 6   production_companies  31747 non-null  object        
 7   production_countries  36536 non-null  object        
 8   release_date          41041 non-null  datetime64[ns]
 9   revenue               41065 non-null  float64       
 10  runtime               40911 non-null  float64       
 11  spoken_languages      38584 non-null  object        
 12  title                 41065 non-null  object        
 13  vote_average          41065 non-null  float64       
 14  vote_count            41065 non-null  float64       
 15  cast                  41065 non-null  object        
 16  crew                  41065 non-null  object        
 17  year                  41041 non-null  float64       
dtypes: datetime64[ns](1), float64(7), int32(1), object(9)
memory usage: 5.8+ MB
In [24]:
movies_clean = movies_clean.dropna(how = 'any', subset = ['genres'])
In [25]:
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 39663 entries, 0 to 45524
Data columns (total 18 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                39663 non-null  float64       
 1   genres                39663 non-null  object        
 2   id                    39663 non-null  int32         
 3   original_language     39663 non-null  object        
 4   original_title        39663 non-null  object        
 5   popularity            39663 non-null  float64       
 6   production_companies  31502 non-null  object        
 7   production_countries  35944 non-null  object        
 8   release_date          39654 non-null  datetime64[ns]
 9   revenue               39663 non-null  float64       
 10  runtime               39546 non-null  float64       
 11  spoken_languages      37734 non-null  object        
 12  title                 39663 non-null  object        
 13  vote_average          39663 non-null  float64       
 14  vote_count            39663 non-null  float64       
 15  cast                  39663 non-null  object        
 16  crew                  39663 non-null  object        
 17  year                  39654 non-null  float64       
dtypes: datetime64[ns](1), float64(7), int32(1), object(9)
memory usage: 5.6+ MB
In [26]:
movies_clean = movies_clean.dropna(how = 'any', subset = ['production_companies','production_countries'])
In [27]:
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 31079 entries, 0 to 45524
Data columns (total 18 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                31079 non-null  float64       
 1   genres                31079 non-null  object        
 2   id                    31079 non-null  int32         
 3   original_language     31079 non-null  object        
 4   original_title        31079 non-null  object        
 5   popularity            31079 non-null  float64       
 6   production_companies  31079 non-null  object        
 7   production_countries  31079 non-null  object        
 8   release_date          31078 non-null  datetime64[ns]
 9   revenue               31079 non-null  float64       
 10  runtime               31029 non-null  float64       
 11  spoken_languages      30568 non-null  object        
 12  title                 31079 non-null  object        
 13  vote_average          31079 non-null  float64       
 14  vote_count            31079 non-null  float64       
 15  cast                  31079 non-null  object        
 16  crew                  31079 non-null  object        
 17  year                  31078 non-null  float64       
dtypes: datetime64[ns](1), float64(7), int32(1), object(9)
memory usage: 4.4+ MB
In [28]:
movies_clean = movies_clean.dropna(how = 'any', subset = ['release_date','runtime','spoken_languages','release_date','year'  ])
In [29]:
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 30528 entries, 0 to 45524
Data columns (total 18 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                30528 non-null  float64       
 1   genres                30528 non-null  object        
 2   id                    30528 non-null  int32         
 3   original_language     30528 non-null  object        
 4   original_title        30528 non-null  object        
 5   popularity            30528 non-null  float64       
 6   production_companies  30528 non-null  object        
 7   production_countries  30528 non-null  object        
 8   release_date          30528 non-null  datetime64[ns]
 9   revenue               30528 non-null  float64       
 10  runtime               30528 non-null  float64       
 11  spoken_languages      30528 non-null  object        
 12  title                 30528 non-null  object        
 13  vote_average          30528 non-null  float64       
 14  vote_count            30528 non-null  float64       
 15  cast                  30528 non-null  object        
 16  crew                  30528 non-null  object        
 17  year                  30528 non-null  float64       
dtypes: datetime64[ns](1), float64(7), int32(1), object(9)
memory usage: 4.3+ MB
In [30]:
movies_clean['year']=movies_clean[['year']].astype(float).astype(int)
In [31]:
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 30528 entries, 0 to 45524
Data columns (total 18 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                30528 non-null  float64       
 1   genres                30528 non-null  object        
 2   id                    30528 non-null  int32         
 3   original_language     30528 non-null  object        
 4   original_title        30528 non-null  object        
 5   popularity            30528 non-null  float64       
 6   production_companies  30528 non-null  object        
 7   production_countries  30528 non-null  object        
 8   release_date          30528 non-null  datetime64[ns]
 9   revenue               30528 non-null  float64       
 10  runtime               30528 non-null  float64       
 11  spoken_languages      30528 non-null  object        
 12  title                 30528 non-null  object        
 13  vote_average          30528 non-null  float64       
 14  vote_count            30528 non-null  float64       
 15  cast                  30528 non-null  object        
 16  crew                  30528 non-null  object        
 17  year                  30528 non-null  int32         
dtypes: datetime64[ns](1), float64(6), int32(2), object(9)
memory usage: 4.2+ MB
In [32]:
def get_director(x):
    for i in x:
        if i['job'] == 'Director':
            return i['name']
    return np.nan
In [33]:
movies_clean['director'] = movies_clean['crew'].apply(ast.literal_eval).apply(get_director)
movies_clean['director']
Out[33]:
0          John Lasseter
1           Joe Johnston
2          Howard Deutch
3        Forest Whitaker
4          Charles Shyer
              ...       
45519           Ben Rock
45520      Aaron Osborne
45521         John Irvin
45523           Lav Diaz
45524     Mark L. Lester
Name: director, Length: 30528, dtype: object
In [34]:
def get_director_gender(x):
    for i in x:
        if i['job'] == 'Director':
            if i['gender'] == 1:
                return "female"
            elif i['gender'] == 2:
                return "male"
            else:
                return "unknown"
    return np.nan
In [35]:
movies_clean['director_gender'] = movies_clean['crew'].apply(ast.literal_eval).apply(get_director_gender)
gender= pd.DataFrame(movies_clean['director_gender'].value_counts())
gender
Out[35]:
director_gender
male 19366
unknown 9955
female 1118
In [46]:
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 30439 entries, 0 to 45524
Data columns (total 20 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                30439 non-null  float64       
 1   genres                30439 non-null  object        
 2   id                    30439 non-null  int32         
 3   original_language     30439 non-null  object        
 4   original_title        30439 non-null  object        
 5   popularity            30439 non-null  float64       
 6   production_companies  30439 non-null  object        
 7   production_countries  30439 non-null  object        
 8   release_date          30439 non-null  datetime64[ns]
 9   revenue               30439 non-null  float64       
 10  runtime               30439 non-null  float64       
 11  spoken_languages      30439 non-null  object        
 12  title                 30439 non-null  object        
 13  vote_average          30439 non-null  float64       
 14  vote_count            30439 non-null  float64       
 15  cast                  30439 non-null  object        
 16  crew                  30439 non-null  object        
 17  year                  30439 non-null  int32         
 18  director              30439 non-null  object        
 19  director_gender       30439 non-null  object        
dtypes: datetime64[ns](1), float64(6), int32(2), object(11)
memory usage: 5.7+ MB
In [37]:
movies_clean = movies_clean.dropna(how = 'any', subset = ['director','director_gender'])
movies_clean.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 30439 entries, 0 to 45524
Data columns (total 20 columns):
 #   Column                Non-Null Count  Dtype         
---  ------                --------------  -----         
 0   budget                30439 non-null  float64       
 1   genres                30439 non-null  object        
 2   id                    30439 non-null  int32         
 3   original_language     30439 non-null  object        
 4   original_title        30439 non-null  object        
 5   popularity            30439 non-null  float64       
 6   production_companies  30439 non-null  object        
 7   production_countries  30439 non-null  object        
 8   release_date          30439 non-null  datetime64[ns]
 9   revenue               30439 non-null  float64       
 10  runtime               30439 non-null  float64       
 11  spoken_languages      30439 non-null  object        
 12  title                 30439 non-null  object        
 13  vote_average          30439 non-null  float64       
 14  vote_count            30439 non-null  float64       
 15  cast                  30439 non-null  object        
 16  crew                  30439 non-null  object        
 17  year                  30439 non-null  int32         
 18  director              30439 non-null  object        
 19  director_gender       30439 non-null  object        
dtypes: datetime64[ns](1), float64(6), int32(2), object(11)
memory usage: 4.6+ MB

1. Kurie filmai daugiausiai uždirbo? Didžiausio biudžeto filmai?¶

In [38]:
stulpeliai = ['title', 'budget']
biudžetas = movies_clean.sort_values('budget', ascending=False)[stulpeliai].set_index('title')
top_10_biudžetas = biudžetas.head(10)

fig, ax = plt.subplots(figsize=(15,5))
sns.set_style('white')
sns.barplot(data=top_10_biudžetas, x=top_10_biudžetas.index, y='budget');
plt.xticks(ha='left', rotation=-20, fontsize=15); plt.yticks(fontsize=15)
plt.xlabel(''); plt.ylabel('USD 100 Million', fontsize=15 );
plt.title('Top 10 didžiausio biudžeto filmai', fontsize=15, weight = 'bold' );

stulpeliai = ['title', 'revenue']
uždarbis = movies_clean.sort_values('revenue', ascending=False)[stulpeliai].set_index('title')
top_10_uždarbis = uždarbis.head(10)

fig, ax = plt.subplots(figsize=(15,5))
sns.set_style('white')
sns.barplot(data=top_10_uždarbis, x=top_10_uždarbis.index, y='revenue');
plt.xticks(ha='left', rotation=-20, fontsize=15); plt.yticks(fontsize=15)
plt.xlabel(''); plt.ylabel('USD Billion', fontsize=15);
plt.title('Top 10 daugiausiai uždirbę filmai', fontsize=15, weight = 'bold' );

pelnas = movies_clean['revenue'] - movies_clean['budget']
pelnas.name = 'profit'
pelnas = movies_clean.join(pelnas)[['title', 'budget', 'revenue', 'profit']].sort_values('profit', ascending=False)
top_10_pelnas = pelnas.head(10).set_index('title')

pelnas_procentais = (top_10_pelnas['profit'] / top_10_pelnas['budget'] * 100)
pelnas_procentais = pelnas_procentais.sort_values(ascending=False).to_frame().rename(columns={0:'Pelnas procentais'})

fig, ax = plt.subplots(figsize=(15,5))
sns.set_style('white')
sns.barplot(data=pelnas_procentais, x=pelnas_procentais.index, y='Pelnas procentais')
plt.xticks(ha='left', rotation=-20, fontsize=15); plt.yticks(fontsize=15)
plt.xlabel(''); plt.ylabel('Pelnas procentais', fontsize=15);
plt.title('Pelnas procentais Top 10 pelningiausių filmų', fontsize=15, weight = 'bold');

Didžiausio biudžeto filmas 'Karibų piratai:ant keistų bangų' Didžiausio uždarbio filmas 'Avataras' Didžiausio biudžeto ir didžiausio uždarbio diagramose matome, kad pasikartoja tik 2 filmai, todėl teigti, kad daugiausiai uždirba didžiausio biudžeto filmai nėra visai teisinga :)
Pelningiausias filmas 'Avataras'. Neįtikėtina, kaip sėkmingi filmai uždirba daugiau nei jų biudžetas. Avataras uždirbo 10 kartų daugiau nei buvo jo biudžetas. „Avengers: Age of Ultron“, šio sąrašo apačioje, beveik 4 kartus viršijo savo biudžetą.

In [132]:
movies_clean[movies_clean['title']== 'Avatar']
Out[132]:
budget genres id original_language original_title popularity production_companies production_countries release_date revenue runtime spoken_languages title vote_average vote_count cast crew year director director_gender
14617 237000000.0 Action 19995 en Avatar 185.070892 Ingenious Film Partners United States of America 2009-12-10 2.787965e+09 162.0 English Avatar 7.2 12114.0 [{'cast_id': 242, 'character': 'Jake Sully', '... [{'credit_id': '52fe48009251416c750aca23', 'de... 2009 James Cameron male

Kurie filmai buvo nuostolingi?¶

In [39]:
top_10_nuostolis = pelnas[pelnas['revenue'] > 0].tail(10).sort_values(['profit', 'revenue']).set_index('title')

plt.style.use('seaborn')
top_10_nuostolis.plot(kind='bar', figsize=(20,6), fontsize=20)
plt.ylabel('USD Billion', fontsize=20); plt.xlabel('')
plt.xticks(rotation=-20, ha='left')
plt.suptitle('Top 10 nuostolingiausių filmų', fontsize=20)
plt.legend(fontsize=15);

Šiek tiek nustebau pamaciusi „The Lone Ranger“ šio sąrašo viršuje. Nors nežiūrėjau šio filmo, žinojau, kad jame vaidina Johnny Depp'as, ir maniau, kad jis buvo sėkmingas. Vadinasi ne visada geras ir žinomas pagrindinis aktorius garantuoja filmui sėkmę.

In [130]:
gifPath = Path("D:\\johnny-depp-the-lone-ranger.gif")
with open(gifPath,'rb') as f:
    display.Image(data=f.read(), format='png')
Out[130]:
In [129]:
movies_clean[movies_clean['title']== 'The Lone Ranger']
Out[129]:
budget genres id original_language original_title popularity production_companies production_countries release_date revenue runtime spoken_languages title vote_average vote_count cast crew year director director_gender
5029 0.0 Adventure 108266 en The Lone Ranger 0.889869 Wrather Productions United States of America 1956-02-25 0.0 86.0 English The Lone Ranger 6.1 7.0 [{'cast_id': 2, 'character': 'The Lone Ranger'... [{'credit_id': '52fe4a94c3a36847f81d632b', 'de... 1956 Stuart Heisler male
21247 255000000.0 Action 57201 en The Lone Ranger 12.729104 Walt Disney Pictures United States of America 2013-07-03 89289910.0 149.0 English The Lone Ranger 5.9 2361.0 [{'cast_id': 4, 'character': 'Tonto', 'credit_... [{'credit_id': '52fe4928c3a36847f818be95', 'de... 2013 Gore Verbinski male

Kurie režisieriai yra sukūrę daugiausiai filmų?¶

In [40]:
plt.figure(figsize=[10,6])
top_dir = movies_clean.director.value_counts().head(10).index
count = movies_clean.director.value_counts().head(10)

plt.barh(top_dir, count, color='#444444')

plt.title('Top 10 režisieriai')
plt.xlabel('Filmų skaičius');

Smagu matyti kad naujoji karta dar nepasivijo senosios kartos, daugiausiai filmu sukurę režisieriai yra senosios kartos kurejai, tokiu filmų kaip 12 įtūžusių vyriškių (1957m.) Kasablanka (1942 m.) Psichopatas(1960 m.). Iš produktyviausių režisierių Top 10 tik 2 režisieriai yra dar gyvi, tai Woody Allen (86metai) ir Werner Herzog (79m) Kitas idomus dalykas yra tas, kad nors ir amerika garseja savo filmu industrija taciau 4 is 10 sio saraso rezisieriu nera amerikieciai: Michaelas Curtizas buvo vengrų kilmės amerikiečių kino režisierius; Alfredas Hičkokas – kino režisierius ir prodiuseris, kilęs iš Didžiosios Britanijos Werneris Herzogas yra vokiečių kino režisierius Ernstas Ingmaras Bergmanas – garsus švedų teatro ir kino režisierius, Taip pat produktyviausių režisierių sarase nera moteru...

Kokio žanro filmų yra kuriama daugiausiai?¶

In [41]:
plt.figure(figsize=[10,6])
zanras = movies_clean.genres.value_counts().index
count = movies_clean.genres.value_counts()
plt.barh(zanras, count, color='#444444')

plt.title('Žanrai')
plt.xlabel('Filmų skaičius');

Kokie žodžiai yra dažniausiai pasikartojantys filmų pavadinimuose?¶

In [123]:
original_corpus = ' '.join(movies_clean['title'])

title_wordcloud = WordCloud(stopwords=STOPWORDS, background_color='black', height=2000, width=4000).generate(original_corpus)
plt.figure(figsize=(20,8))
plt.imshow(title_wordcloud)
plt.axis('off')
plt.show()
Out[123]:
<Figure size 1440x576 with 0 Axes>
Out[123]:
<matplotlib.image.AxesImage at 0x1c7a9d5a430>
Out[123]:
(-0.5, 3999.5, 1999.5, -0.5)

Kurie filmai TMDB bendruomenės įvertinti geriausiai?¶

In [118]:
movies_clean[movies_clean['vote_count'] > 5000][['title', 'vote_average', 'vote_count' ,'year', 'popularity']].sort_values('vote_average', ascending=False).head(10)
Out[118]:
title vote_average vote_count year popularity
314 The Shawshank Redemption 8.5 8358.0 1994 51.645403
837 The Godfather 8.5 6024.0 1972 41.109264
292 Pulp Fiction 8.3 8670.0 1994 140.950236
12525 The Dark Knight 8.3 12269.0 2008 123.167259
2854 Fight Club 8.3 9678.0 1999 63.869599
18537 The Intouchables 8.2 5410.0 2011 16.086919
351 Forrest Gump 8.2 8147.0 1994 48.307194
1163 The Empire Strikes Back 8.2 5998.0 1980 19.470959
256 Star Wars 8.1 6778.0 1977 42.149697
46 Se7en 8.1 5915.0 1995 18.457430

„Shawshank Redemption“ ir „The Godfather“ yra du geriausiai TMDB bendruomenės įvertinti filmai. Įdomu tai, kad jie taip pat yra 2 geriausi filmai IMDB 250 geriausių filmų sąraše. Taip pat, įdomu tai, kad 8 iš 10 geriausiai įvertintų filmų buvo išleisti dar prieš 2000m.

In [44]:
vote_avg=np.arange(0,10+0.5,0.5)
plt.hist(data=movies_clean, x='vote_average',bins=vote_avg);
plt.title('Balsu vidurkis');

Kaip kito filmų ilgis bėgant metams?¶

In [97]:
movies_clean.plot.scatter(x='year', y = 'runtime', color=base_color)
Out[97]:
<AxesSubplot:xlabel='year', ylabel='runtime'>
In [99]:
movies_clean[movies_clean.runtime > 800]
Out[99]:
budget genres id original_language original_title popularity production_companies production_countries release_date revenue runtime spoken_languages title vote_average vote_count cast crew year director director_gender
13824 0.0 Drama 45560 de Berlin Alexanderplatz 2.255785 Bavaria Film Germany 1980-08-28 0.0 931.0 Deutsch Berlin Alexanderplatz 8.4 5.0 [{'cast_id': 2, 'character': 'Franz Biberkopf'... [{'credit_id': '5653d1dac3a36850fc001c27', 'de... 1980 Rainer Werner Fassbinder male
19230 0.0 Documentary 208988 en The War 0.487489 PBS United States of America 2007-09-23 0.0 874.0 English The War 5.3 3.0 [{'cast_id': 3, 'character': 'Narrator', 'cred... [{'credit_id': '52fe4d58c3a368484e1e43a5', 'de... 2007 Ken Burns male
26763 0.0 Science Fiction 150004 en Taken 10.628230 DreamWorks SKG United States of America 2002-12-02 0.0 877.0 English Taken 7.4 76.0 [{'cast_id': 1, 'character': 'Allie Keys', 'cr... [{'credit_id': '54177f100e0a2637ef0002e0', 'de... 2002 John Fawcett male
34744 0.0 History 293603 en The Roosevelts: An Intimate History 0.824032 Florentine Films United States of America 2014-09-15 0.0 840.0 English The Roosevelts: An Intimate History 8.3 6.0 [{'cast_id': 0, 'character': 'Eleanor Roosevel... [{'credit_id': '541f65a70e0a261794000b7f', 'de... 2014 Ken Burns male
In [ ]:

Kino pramone valdo vyrai?¶

In [107]:
gender= pd.DataFrame(movies_clean['director_gender'].value_counts())
gender
Out[107]:
director_gender
male 19366
unknown 9955
female 1118
In [45]:
sns.catplot(x="year", hue="director_gender", col="director_gender", data=movies_clean, kind="strip", legend=True);

Nors ir neturime visų režisierių lyties duomenų, gana akivaizdu, kad moterų visame pasaulyje filmų kūrimo pramonėje yra mažuma. Tačiau galime pastebėti kintantį augimą 2000 m.

In [106]:
movies_clean[movies_clean['director_gender']== 'female'][['title', 'vote_count', 'vote_average', 'director_gender', 'director', 'popularity','year']].sort_values('popularity', ascending=False).head(10)
Out[106]:
title vote_count vote_average director_gender director popularity year
33435 Wonder Woman 5025.0 7.2 female Patty Jenkins 294.337037 2017
44338 The Bad Batch 160.0 5.3 female Ana Lily Amirpour 78.807200 2017
43710 The Beguiled 279.0 5.8 female Sofia Coppola 36.260510 2017
36298 Me Before You 2674.0 7.6 female Thea Sharrock 34.347590 2016
2466 The Matrix 9079.0 7.9 female Lana Wachowski 33.366332 1999
27559 Fifty Shades of Grey 3350.0 5.2 female Sam Taylor-Johnson 33.068431 2015
13177 Twilight 3688.0 5.8 female Catherine Hardwicke 31.282029 2008
23962 Jupiter Ascending 2816.0 5.2 female Lilly Wachowski 21.463465 2015
41572 Underworld: Blood Wars 1619.0 5.2 female Anna Foerster 20.029760 2016
8215 Shark Tale 1612.0 5.8 female Vicky Jenson 17.999273 2004
In [100]:
gifPath = Path("D:\\giphy (5).gif")
with open(gifPath,'rb') as f:
    display.Image(data=f.read(), format='png')
Out[100]:
In [ ]:
In [ ]: