Skip to content

Latest commit

 

History

History
2892 lines (2404 loc) · 49.2 KB

README.md

File metadata and controls

2892 lines (2404 loc) · 49.2 KB
Raw

LetterBoxd-Movie-Scrapping

This project extracts movie data from LetterBoxd and performs Exploratory Data Analysis to understand trends in movies and different movie genres.

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
import warnings
warnings.filterwarnings('ignore')
df = pd.read_json("feb2021.json")
df.head()
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
title year duration genre rating language viewers
0 Parasite 2019.0 133 [comedy, drama, thriller] 4.6 Korean Weighted average of 4.60 based on 705,993 ratings
1 Call Me by Your Name 2017.0 132 [drama, romance] 4.1 English Weighted average of 4.09 based on 355,937 ratings
2 Hereditary 2018.0 127 [thriller, horror, mystery] 4.0 English Weighted average of 3.99 based on 365,955 ratings
3 Spirited Away 2001.0 125 [fantasy, animation, family] 4.5 Japanese Weighted average of 4.47 based on 386,917 ratings
4 Moonlight 2016.0 111 [drama] 4.2 English Weighted average of 4.20 based on 347,384 ratings 
df.shape
(94051, 7)

df.columns
Index(['title', 'year', 'duration', 'genre', 'rating', 'language', 'viewers'], dtype='object')

df.isnull().sum()
title           0
year         3379
duration        0
genre           0
rating      70251
language      586
viewers     70251
dtype: int64

df = df.loc[~(df.year.isnull())]
df = df.loc[~(df.rating.isnull())]
df = df.loc[~(df.language.isnull())]
df.isnull().sum()
title       0
year        0
duration    0
genre       0
rating      0
language    0
viewers     0
dtype: int64

df.columns
Index(['title', 'year', 'duration', 'genre', 'rating', 'language', 'viewers'], dtype='object')

df.year = df.year.astype('int')
df.year = df.year.astype('category')
df.year.value_counts()
2019    1247
2018    1218
2017    1196
2016    1090
2015    1003
        ... 
1895       8
1894       5
1893       1
1888       1
1886       1
Name: year, Length: 131, dtype: int64

df.loc[(df.duration=="")]
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
title year duration genre rating language viewers
715 The King's Man 2021 [action, adventure, comedy] 3.3 English Weighted average of 3.29 based on 50 ratings
969 The Matrix Resurrections 2021 [action, adventure, science fiction] 3.8 English Weighted average of 3.76 based on 97 ratings
1958 Coming 2 America 2021 [family, comedy] 3.3 English Weighted average of 3.27 based on 57 ratings
2270 Minions: The Rise of Gru 2021 [family, action, adventure, comedy, animation] 2.6 English Weighted average of 2.65 based on 237 ratings
2891 Hotel Transylvania 4 2021 [adventure, animation, family, fantasy] 3.3 English Weighted average of 3.30 based on 113 ratings
... ... ... ... ... ... ... ...
41910 Murdered 2012 [crime] 3.0 French Weighted average of 3.05 based on 30 ratings
42148 New Moon 2010 [action] 2.9 English Weighted average of 2.90 based on 31 ratings
42179 Uthaman 2001 [] 3.1 Malayalam Weighted average of 3.08 based on 32 ratings
42906 Daniël Arends: Joko 79 2008 [comedy] 3.4 Dutch Weighted average of 3.42 based on 30 ratings
45660 Simiocracia (Crónica de la gran resaca económica) 2012 [comedy, animation, documentary] 3.3 Spanish Weighted average of 3.25 based on 32 ratings

851 rows × 7 columns

len(df.loc[(df.duration=="")])
851

df =df.loc[~(df.duration=="")]
df.duration = df.duration.astype("int")
df.duration.describe()
count    22793.000000
mean        84.253631
std         69.173581
min          1.000000
25%         65.000000
50%         90.000000
75%        103.000000
max       3708.000000
Name: duration, dtype: float64

df.columns
Index(['title', 'year', 'duration', 'genre', 'rating', 'language', 'viewers'], dtype='object')

df.genre.head()
0       [comedy, drama, thriller]
1                [drama, romance]
2     [thriller, horror, mystery]
3    [fantasy, animation, family]
4                         [drama]
Name: genre, dtype: object

myGenre = set()
df.shape
(22793, 7)

df = df.reset_index()
df.drop("index", axis=1, inplace=True)
for i in range(0,len(df)):
    for movGen in df.loc[i].genre:
        myGenre.add(movGen)
df.head()
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
title year duration genre rating language viewers
0 Parasite 2019 133 [comedy, drama, thriller] 4.6 Korean Weighted average of 4.60 based on 705,993 ratings
1 Call Me by Your Name 2017 132 [drama, romance] 4.1 English Weighted average of 4.09 based on 355,937 ratings
2 Hereditary 2018 127 [thriller, horror, mystery] 4.0 English Weighted average of 3.99 based on 365,955 ratings
3 Spirited Away 2001 125 [fantasy, animation, family] 4.5 Japanese Weighted average of 4.47 based on 386,917 ratings
4 Moonlight 2016 111 [drama] 4.2 English Weighted average of 4.20 based on 347,384 ratings 
myGenre
{'action',
 'adventure',
 'animation',
 'comedy',
 'crime',
 'documentary',
 'drama',
 'family',
 'fantasy',
 'history',
 'horror',
 'music',
 'mystery',
 'romance',
 'science fiction',
 'thriller',
 'tv movie',
 'war',
 'western'}

len(myGenre)
19

for col in myGenre:
    df[col] = 0
df.head()
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
title year duration genre rating language viewers thriller tv movie mystery ... action music history comedy science fiction romance crime animation war adventure
0 Parasite 2019 133 [comedy, drama, thriller] 4.6 Korean Weighted average of 4.60 based on 705,993 ratings 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
1 Call Me by Your Name 2017 132 [drama, romance] 4.1 English Weighted average of 4.09 based on 355,937 ratings 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
2 Hereditary 2018 127 [thriller, horror, mystery] 4.0 English Weighted average of 3.99 based on 365,955 ratings 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
3 Spirited Away 2001 125 [fantasy, animation, family] 4.5 Japanese Weighted average of 4.47 based on 386,917 ratings 0 0 0 ... 0 0 0 0 0 0 0 0 0 0
4 Moonlight 2016 111 [drama] 4.2 English Weighted average of 4.20 based on 347,384 ratings 0 0 0 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 26 columns

for i in range(0,len(df)):
    for x in df.loc[i]["genre"]:
        df.at[i, x]=1
df.head()
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
title year duration genre rating language viewers thriller tv movie mystery ... action music history comedy science fiction romance crime animation war adventure
0 Parasite 2019 133 [comedy, drama, thriller] 4.6 Korean Weighted average of 4.60 based on 705,993 ratings 1 0 0 ... 0 0 0 1 0 0 0 0 0 0
1 Call Me by Your Name 2017 132 [drama, romance] 4.1 English Weighted average of 4.09 based on 355,937 ratings 0 0 0 ... 0 0 0 0 0 1 0 0 0 0
2 Hereditary 2018 127 [thriller, horror, mystery] 4.0 English Weighted average of 3.99 based on 365,955 ratings 1 0 1 ... 0 0 0 0 0 0 0 0 0 0
3 Spirited Away 2001 125 [fantasy, animation, family] 4.5 Japanese Weighted average of 4.47 based on 386,917 ratings 0 0 0 ... 0 0 0 0 0 0 0 1 0 0
4 Moonlight 2016 111 [drama] 4.2 English Weighted average of 4.20 based on 347,384 ratings 0 0 0 ... 0 0 0 0 0 0 0 0 0 0

5 rows × 26 columns



df.columns


Index(['title', 'year', 'duration', 'genre', 'rating', 'language', 'viewers',
       'thriller', 'tv movie', 'mystery', 'documentary', 'western', 'family',
       'horror', 'drama', 'fantasy', 'action', 'music', 'history', 'comedy',
       'science fiction', 'romance', 'crime', 'animation', 'war', 'adventure'],
      dtype='object')



df.rating = df.rating.astype("float")


df.rating.describe()


count    22793.000000
mean         3.137270
std          0.429866
min          1.000000
25%          2.900000
50%          3.200000
75%          3.400000
max          4.600000
Name: rating, dtype: float64



df.language.value_counts().head(10)


English       13344
French         1276
Japanese       1078
Spanish         952
German          644
Portuguese      588
Italian         547
Chinese         340
Hindi           315
Korean          303
Name: language, dtype: int64



df.viewers.head()


0    Weighted average of 4.60 based on 705,993 ratings
1    Weighted average of 4.09 based on 355,937 ratings
2    Weighted average of 3.99 based on 365,955 ratings
3    Weighted average of 4.47 based on 386,917 ratings
4    Weighted average of 4.20 based on 347,384 ratings
Name: viewers, dtype: object



df.viewers = df.viewers.apply(lambda x: int(x.split()[-2].replace(",","")))


df.viewers.describe()


count     22793.000000
mean       2443.237705
std       18353.907074
min          30.000000
25%          50.000000
50%         102.000000
75%         324.000000
max      705993.000000
Name: viewers, dtype: float64



Analysis


fig = plt.figure(figsize=(20,6))
axis1 = fig.add_subplot(1,2,1)
axis2 = fig.add_subplot(1,2,2)
sns.boxplot(x = "rating", data=df, ax=axis1)
sns.boxplot(x = "duration", data=df, ax=axis2)
plt.show()


png


sns.boxplot(x = "viewers", data=df)
plt.show()


png


df.columns


Index(['title', 'year', 'duration', 'genre', 'rating', 'language', 'viewers',
       'thriller', 'tv movie', 'mystery', 'documentary', 'western', 'family',
       'horror', 'drama', 'fantasy', 'action', 'music', 'history', 'comedy',
       'science fiction', 'romance', 'crime', 'animation', 'war', 'adventure'],
      dtype='object')



df.columns[0:7]


Index(['title', 'year', 'duration', 'genre', 'rating', 'language', 'viewers'], dtype='object')



df.columns[7:]


Index(['thriller', 'tv movie', 'mystery', 'documentary', 'western', 'family',
       'horror', 'drama', 'fantasy', 'action', 'music', 'history', 'comedy',
       'science fiction', 'romance', 'crime', 'animation', 'war', 'adventure'],
      dtype='object')



pd.melt(df, id_vars = df.columns[0:7], value_vars=df.columns[7:])



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      title
      year
      duration
      genre
      rating
      language
      viewers
      variable
      value
    

  
  
    

      0
      Parasite
      2019
      133
      [comedy, drama, thriller]
      4.6
      Korean
      705993
      thriller
      1
    

    

      1
      Call Me by Your Name
      2017
      132
      [drama, romance]
      4.1
      English
      355937
      thriller
      0
    

    

      2
      Hereditary
      2018
      127
      [thriller, horror, mystery]
      4.0
      English
      365955
      thriller
      1
    

    

      3
      Spirited Away
      2001
      125
      [fantasy, animation, family]
      4.5
      Japanese
      386917
      thriller
      0
    

    

      4
      Moonlight
      2016
      111
      [drama]
      4.2
      English
      347384
      thriller
      0
    

    

      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
    

    

      433062
      Streets of Rio
      2007
      97
      [drama]
      3.1
      English
      30
      adventure
      0
    

    

      433063
      Whose Dominion?
      2017
      10
      [history, documentary]
      2.9
      English
      41
      adventure
      0
    

    

      433064
      Lazania
      2019
      90
      [comedy]
      2.4
      Persian (Farsi)
      33
      adventure
      0
    

    

      433065
      Prisoners
      2019
      101
      [comedy]
      2.5
      Persian (Farsi)
      30
      adventure
      0
    

    

      433066
      Muraren
      2002
      95
      []
      3.5
      English
      33
      adventure
      0
    

  



433067 rows × 9 columns




df_pivot = pd.melt(df, id_vars = df.columns[0:7], value_vars=df.columns[7:])


df_pivot



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      title
      year
      duration
      genre
      rating
      language
      viewers
      variable
      value
    

  
  
    

      0
      Parasite
      2019
      133
      [comedy, drama, thriller]
      4.6
      Korean
      705993
      thriller
      1
    

    

      1
      Call Me by Your Name
      2017
      132
      [drama, romance]
      4.1
      English
      355937
      thriller
      0
    

    

      2
      Hereditary
      2018
      127
      [thriller, horror, mystery]
      4.0
      English
      365955
      thriller
      1
    

    

      3
      Spirited Away
      2001
      125
      [fantasy, animation, family]
      4.5
      Japanese
      386917
      thriller
      0
    

    

      4
      Moonlight
      2016
      111
      [drama]
      4.2
      English
      347384
      thriller
      0
    

    

      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
      ...
    

    

      433062
      Streets of Rio
      2007
      97
      [drama]
      3.1
      English
      30
      adventure
      0
    

    

      433063
      Whose Dominion?
      2017
      10
      [history, documentary]
      2.9
      English
      41
      adventure
      0
    

    

      433064
      Lazania
      2019
      90
      [comedy]
      2.4
      Persian (Farsi)
      33
      adventure
      0
    

    

      433065
      Prisoners
      2019
      101
      [comedy]
      2.5
      Persian (Farsi)
      30
      adventure
      0
    

    

      433066
      Muraren
      2002
      95
      []
      3.5
      English
      33
      adventure
      0
    

  



433067 rows × 9 columns




df_pivot = df_pivot.loc[df_pivot.value==1]


df_pivot.head()



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      title
      year
      duration
      genre
      rating
      language
      viewers
      variable
      value
    

  
  
    

      0
      Parasite
      2019
      133
      [comedy, drama, thriller]
      4.6
      Korean
      705993
      thriller
      1
    

    

      2
      Hereditary
      2018
      127
      [thriller, horror, mystery]
      4.0
      English
      365955
      thriller
      1
    

    

      7
      1917
      2019
      119
      [thriller, war, drama, action]
      4.1
      English
      358009
      thriller
      1
    

    

      8
      Gone Girl
      2014
      149
      [drama, thriller, mystery]
      4.0
      English
      445648
      thriller
      1
    

    

      10
      The Lighthouse
      2019
      109
      [thriller, fantasy, drama, horror]
      4.1
      English
      304207
      thriller
      1
    

  





df_pivot.shape


(39384, 9)



df_pivot.loc[df_pivot.title=="Soul"]



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      title
      year
      duration
      genre
      rating
      language
      viewers
      variable
      value
    

  
  
    

      113982
      Soul
      2020
      102
      [family, drama, music, comedy, animation, fant...
      4.1
      English
      281921
      family
      1
    

    

      159568
      Soul
      2020
      102
      [family, drama, music, comedy, animation, fant...
      4.1
      English
      281921
      drama
      1
    

    

      182361
      Soul
      2020
      102
      [family, drama, music, comedy, animation, fant...
      4.1
      English
      281921
      fantasy
      1
    

    

      227947
      Soul
      2020
      102
      [family, drama, music, comedy, animation, fant...
      4.1
      English
      281921
      music
      1
    

    

      273533
      Soul
      2020
      102
      [family, drama, music, comedy, animation, fant...
      4.1
      English
      281921
      comedy
      1
    

    

      364705
      Soul
      2020
      102
      [family, drama, music, comedy, animation, fant...
      4.1
      English
      281921
      animation
      1
    

  





df_pivot.drop(["genre", "value"], axis=1, inplace=True)


df_pivot.head()



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      title
      year
      duration
      rating
      language
      viewers
      variable
    

  
  
    

      0
      Parasite
      2019
      133
      4.6
      Korean
      705993
      thriller
    

    

      2
      Hereditary
      2018
      127
      4.0
      English
      365955
      thriller
    

    

      7
      1917
      2019
      119
      4.1
      English
      358009
      thriller
    

    

      8
      Gone Girl
      2014
      149
      4.0
      English
      445648
      thriller
    

    

      10
      The Lighthouse
      2019
      109
      4.1
      English
      304207
      thriller
    

  





df_pivot.rename(columns={"variable": "genre"}, inplace=True)


df = df_pivot


def annotate_graph_both(ax):
    total_height = 0
    for bar in ax.patches:
        total_height=total_height+bar.get_height()
        
#     for bar in ax.patches: 
#         ax.annotate(format((bar.get_height()/total_height )* 100, '.1f')+" %",  
#                    (bar.get_x() + bar.get_width() / 2,  bar.get_height()/2), 
#                    ha='center', va='center', 
#                    size=10, xytext=(0, 8), 
#                    textcoords='offset points')
        
    for bar in ax.patches: 
        ax.annotate(format((bar.get_height()), '.2f'),  
                   (bar.get_x() + bar.get_width() / 2,  bar.get_height()), 
                   ha='center', va='center', 
                   size=10, xytext=(0, 8), 
                   textcoords='offset points')
    return ax


grouped_rating_median = df.loc[:,['genre', 'rating']].groupby(['genre']).median().sort_values(by='rating')
grouped_rating_median = pd.DataFrame(grouped_rating_median)
grouped_rating_median = grouped_rating_median.reset_index()
grouped_rating_median



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      genre
      rating
    

  
  
    

      0
      horror
      2.8
    

    

      1
      thriller
      2.9
    

    

      2
      science fiction
      2.9
    

    

      3
      action
      3.0
    

    

      4
      adventure
      3.0
    

    

      5
      tv movie
      3.0
    

    

      6
      romance
      3.1
    

    

      7
      mystery
      3.1
    

    

      8
      western
      3.1
    

    

      9
      family
      3.1
    

    

      10
      crime
      3.1
    

    

      11
      comedy
      3.1
    

    

      12
      fantasy
      3.1
    

    

      13
      war
      3.2
    

    

      14
      drama
      3.2
    

    

      15
      animation
      3.3
    

    

      16
      history
      3.3
    

    

      17
      music
      3.4
    

    

      18
      documentary
      3.4
    

  





grouped_rating_mean = df.loc[:,['genre', 'rating']].groupby(['genre']).mean().sort_values(by='rating')
grouped_rating_mean = pd.DataFrame(grouped_rating_mean)
grouped_rating_mean = grouped_rating_mean.reset_index()
grouped_rating_mean



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      genre
      rating
    

  
  
    

      0
      horror
      2.800131
    

    

      1
      science fiction
      2.889947
    

    

      2
      thriller
      2.918565
    

    

      3
      action
      2.938684
    

    

      4
      adventure
      2.986467
    

    

      5
      family
      2.991188
    

    

      6
      tv movie
      3.015921
    

    

      7
      fantasy
      3.041466
    

    

      8
      mystery
      3.053564
    

    

      9
      comedy
      3.057474
    

    

      10
      romance
      3.078985
    

    

      11
      crime
      3.104104
    

    

      12
      western
      3.125157
    

    

      13
      drama
      3.192025
    

    

      14
      war
      3.228829
    

    

      15
      animation
      3.239053
    

    

      16
      history
      3.295238
    

    

      17
      music
      3.354450
    

    

      18
      documentary
      3.410717
    

  





fig = plt.figure(figsize=(20,8), dpi=600)
ax1 = fig.add_subplot(2,1,1)
ax2 = fig.add_subplot(2,1,2)

sns.despine()
fig1=sns.barplot(x="genre",y="rating",data=grouped_rating_median, order=grouped_rating_median.genre, ax=ax1)
annotate_graph_both(fig1)
fig1.set_ylabel("median rating")

sns.despine()
fig2=sns.barplot(x="genre",y="rating",data=grouped_rating_mean, order=grouped_rating_mean.genre, ax=ax2)
annotate_graph_both(fig2)
fig2.set_ylabel("avg. rating")
plt.show()


png


OBSERVATION


COMEDY IS HEAVILY AFFECTED BY OUTLIERS


fig = plt.figure(figsize=(20,6))
sns.boxplot(x = "genre", y="rating" ,data=df, order=grouped.index)
plt.show()


png


fig = plt.figure(figsize=(20,6))
sns.violinplot(x = "genre", y="rating" ,data=df, order=grouped_rating_median.index)
plt.show()


png


#--------------------------------------------- DURATOIN -------------------------------------------------#


grouped_duration_median = df.loc[:,['genre', 'duration']].groupby(['genre']).median().sort_values(by='duration')
grouped_duration_median = pd.DataFrame(grouped_duration_median)
grouped_duration_median = grouped_duration_median.reset_index()
grouped_duration_median



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      genre
      duration
    

  
  
    

      0
      animation
      11
    

    

      1
      documentary
      78
    

    

      2
      horror
      88
    

    

      3
      tv movie
      89
    

    

      4
      family
      89
    

    

      5
      fantasy
      89
    

    

      6
      science fiction
      89
    

    

      7
      comedy
      91
    

    

      8
      music
      92
    

    

      9
      western
      93
    

    

      10
      mystery
      93
    

    

      11
      adventure
      94
    

    

      12
      crime
      95
    

    

      13
      thriller
      95
    

    

      14
      action
      96
    

    

      15
      drama
      97
    

    

      16
      romance
      99
    

    

      17
      war
      102
    

    

      18
      history
      106
    

  





grouped_duration_mean = df.loc[:,['genre', 'duration']].groupby(['genre']).mean().sort_values(by='duration')
grouped_duration_mean = pd.DataFrame(grouped_duration_mean)
grouped_duration_mean = grouped_duration_mean.reset_index()
grouped_duration_mean



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      genre
      duration
    

  
  
    

      0
      animation
      49.370629
    

    

      1
      documentary
      76.599204
    

    

      2
      fantasy
      77.324976
    

    

      3
      horror
      79.642826
    

    

      4
      family
      80.514555
    

    

      5
      science fiction
      81.720459
    

    

      6
      comedy
      85.384798
    

    

      7
      tv movie
      88.597368
    

    

      8
      western
      89.503145
    

    

      9
      music
      90.194379
    

    

      10
      adventure
      94.015736
    

    

      11
      thriller
      96.223235
    

    

      12
      mystery
      99.488470
    

    

      13
      romance
      99.847783
    

    

      14
      action
      100.359537
    

    

      15
      drama
      101.025998
    

    

      16
      crime
      101.373134
    

    

      17
      war
      103.414414
    

    

      18
      history
      112.403880
    

  





fig = plt.figure(figsize=(20,8), dpi=600)
ax1 = fig.add_subplot(2,1,1)
ax2 = fig.add_subplot(2,1,2)

sns.despine()
fig1=sns.barplot(x="genre",y="duration",data=grouped_duration_median, order=grouped_duration_median.genre, ax=ax1)
annotate_graph_both(fig1)
fig1.set_ylabel("median duration")

sns.despine()
fig2=sns.barplot(x="genre",y="duration",data=grouped_duration_mean, order=grouped_duration_mean.genre, ax=ax2)
annotate_graph_both(fig2)
fig2.set_ylabel("avg. duration")
plt.show()


png


fig = plt.figure(figsize=(20,6))
sns.boxplot(x = "genre", y="duration" ,data=df.loc[~(df.duration>300)], order=grouped_rating_median.genre)
plt.show()


png


fig = plt.figure(figsize=(20,6))
sns.violinplot(x = "genre", y="duration" ,data=df.loc[~(df.duration>300)], order=grouped_rating_median.genre)
plt.show()


png


df.columns


Index(['title', 'year', 'duration', 'rating', 'language', 'viewers', 'genre'], dtype='object')



fig = plt.figure(figsize=(20,6), dpi=300)
# axis1 = fig.add_subplot(1,2,1)
# axis2 = fig.add_subplot(1,2,2)
sns.scatterplot(hue = "genre", y="rating",x="year", size="duration",data=df, sizes=(10,250))
# sns.boxplot(x = "duration", data=df, ax=axis2)
plt.show()


png


#----------------------------------------------------------------------------------------------------------------#


grouped_viewers_mean = df.loc[:,['genre', 'viewers']].groupby(['genre']).mean().sort_values(by='viewers')
grouped_viewers_mean = pd.DataFrame(grouped_viewers_mean)
grouped_viewers_mean = grouped_viewers_mean.reset_index()
grouped_viewers_mean



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      genre
      viewers
    

  
  
    

      0
      documentary
      378.595886
    

    

      1
      tv movie
      652.826316
    

    

      2
      western
      1712.000000
    

    

      3
      music
      2240.846604
    

    

      4
      horror
      2624.518535
    

    

      5
      comedy
      2695.007110
    

    

      6
      drama
      3087.450755
    

    

      7
      war
      3091.150901
    

    

      8
      animation
      3097.246907
    

    

      9
      romance
      3221.518977
    

    

      10
      crime
      3540.656716
    

    

      11
      history
      3729.504409
    

    

      12
      thriller
      4492.473424
    

    

      13
      family
      5471.936271
    

    

      14
      mystery
      5710.663522
    

    

      15
      action
      6085.403806
    

    

      16
      fantasy
      7251.688525
    

    

      17
      science fiction
      8458.423280
    

    

      18
      adventure
      10093.222659
    

  





grouped_viewers_median = df.loc[:,['genre', 'viewers']].groupby(['genre']).median().sort_values(by='viewers')
grouped_viewers_median = pd.DataFrame(grouped_viewers_median)
grouped_viewers_median = grouped_viewers_median.reset_index()
grouped_viewers_median



<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}


</style>


  
    

      
      genre
      viewers
    

  
  
    

      0
      documentary
      80.5
    

    

      1
      music
      99.0
    

    

      2
      tv movie
      101.0
    

    

      3
      animation
      109.0
    

    

      4
      western
      125.5
    

    

      5
      drama
      127.0
    

    

      6
      comedy
      129.0
    

    

      7
      war
      142.0
    

    

      8
      romance
      143.0
    

    

      9
      history
      145.0
    

    

      10
      horror
      146.0
    

    

      11
      family
      153.0
    

    

      12
      mystery
      157.5
    

    

      13
      action
      165.0
    

    

      14
      thriller
      166.5
    

    

      15
      science fiction
      171.0
    

    

      16
      crime
      171.5
    

    

      17
      adventure
      179.0
    

    

      18
      fantasy
      191.0
    

  





fig = plt.figure(figsize=(20,8), dpi=600)
ax1 = fig.add_subplot(2,1,1)
ax2 = fig.add_subplot(2,1,2)

sns.despine()
fig1=sns.barplot(x="genre",y="viewers",data=grouped_viewers_median, order=grouped_viewers_median.genre, ax=ax1)
annotate_graph_both(fig1)
fig1.set_ylabel("median viewers")

sns.despine()
fig2=sns.barplot(x="genre",y="viewers",data=grouped_viewers_mean, order=grouped_viewers_mean.genre, ax=ax2)
annotate_graph_both(fig2)
fig2.set_ylabel("avg. viewers")
plt.show()


png


Observation


Documentries are highest rated but least viewed. Appraently, "documentries" enojoy an eclectic group who more often than not like what they see and rate it very highly.