Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.

html

Netflix Exploratory Data Analysis - Pavithran S

In [1]: import numpy as np

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import warnings
warnings.filterwarnings("ignore")

In [2]: # using pd.read_csv we can load a csv file.

df = pd.read_csv("netflix.csv")
df.head(5)

Out[2]: show_id type title director cast country date_added release_year rating duration listed_in description

As her
Dick father nears
Kirsten United September
0 s1 Movie Johnson Is NaN 2020 PG-13 90 min Documentaries the end of
Johnson States 25, 2021
Dead his life,
filmm...

Ama
After
Qamata,
International crossing
Khosi
TV Blood & South September TV- 2 TV Shows, TV paths at a
1 s2 NaN Ngema, 2021
Show Water Africa 24, 2021 MA Seasons Dramas, TV party, a
Gail
Mysteries Cape Town
Mabalane,
t...
Thaban...

Sami
Bouajila, Crime TV To protect
Tracy Shows, his family
TV Julien September TV-
2 s3 Ganglands Gotoas, NaN 2021 1 Season International from a
Show Leclercq 24, 2021 MA
Samuel TV Shows, TV powerful
Jouy, Act... drug lor...
Nabi...

Feuds,
Jailbirds flirtations
TV September TV- Docuseries,
3 s4 New NaN NaN NaN 2021 1 Season and toilet
Show 24, 2021 MA Reality TV
Orleans talk go
down amo...

Mayur
In a city of
More, International
coaching
TV Kota Jitendra September TV- 2 TV Shows,
4 s5 NaN India 2021 centers
Show Factory Kumar, 24, 2021 MA Seasons Romantic TV
known to
Ranjan Raj, Shows, TV ...
train I...
Alam K...

Features of the dataset:

• Show_id: Unique ID for every Movie / Tv Show
• Type: Identifier - A Movie or TV Show
• Title: Title of the Movie / Tv Show
• Director: Director of the Movie
• Cast: Actors involved in the movie/show
• Country: Country where the movie/show was produced
• Date_added: Date it was added on Netflix
• Release_year: Actual Release year of the movie/show
• Rating: TV Rating of the movie/show
• Duration: Total Duration - in minutes or number of seasons
• Listed_in: Genre
• Description: The summary description

FINDING SHAPE OF THE DATAFRAME

In [3]: # using df.shape we come to know the given dataset have 8807 rows and 12 columns
df.shape

Out[3]: (8807, 12)

1 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

In [4]: # using df.ndim we come to know the given dataset is a 2dimnesion

df.ndim

Out[4]: 2

FINDING DATA TYPES OF EACH COLUMN

In [5]: # using df.info we come to know the data type of each series
# most of the series is in object datatype and release year is in int datatype format
# we found that the columns = (director, cast, country, date_added, rating, duration) have null (or) missing values
df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8807 entries, 0 to 8806
Data columns (total 12 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 show_id 8807 non-null object
1 type 8807 non-null object
2 title 8807 non-null object
3 director 6173 non-null object
4 cast 7982 non-null object
5 country 7976 non-null object
6 date_added 8797 non-null object
7 release_year 8807 non-null int64
8 rating 8803 non-null object
9 duration 8804 non-null object
10 listed_in 8807 non-null object
11 description 8807 non-null object
dtypes: int64(1), object(11)
memory usage: 825.8+ KB

In [6]: df["type"].value_counts()

Out[6]: type
Movie 6131
TV Show 2676
Name: count, dtype: int64

We got 6131 Movies and 2676 Tv Shows in the given dataset

In [7]: # TV Shows Vs Movies in a Graphical Representation
df_type = df["type"].value_counts()
x = df_type.index
y = df_type
fig = plt.figure (figsize= (10,10))
plt.subplot (2,2,1)
plt.pie(y,explode = (0.03, 0.08), labels = x, colors= ["purple", "orange"], autopct = "%.1f%%")
plt.title("Pie Chart")

plt.subplot (2,2,2)
height = plt.bar(x,y, color = ["purple" ,"orange"])
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Movies & TV Shows")
plt.xlabel("Type")
plt.ylabel("Count")
plt.suptitle("TV Shows Vs Movies " , fontsize = 24)
plt.show()

2 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

UNNESTING THE VALUES

In [8]: # Before Exploratry Data Analysis we have to clean the given data set because it contains null values and nested values.
# first we will unest the nested values
# we have nested values in following columns (director, cast, country, listed_in)
# we are going to unnest director column
df_director = df[["title", "director"]]
df_director["director_unnested"] = df_director["director"].apply(lambda a: str(a).split(", "))
df_director = df_director.explode("director_unnested")
df_director.head(5)

Out[8]: title director director_unnested

0 Dick Johnson Is Dead Kirsten Johnson Kirsten Johnson

1 Blood & Water NaN nan

2 Ganglands Julien Leclercq Julien Leclercq

3 Jailbirds New Orleans NaN nan

4 Kota Factory NaN nan

In [9]: # we have successfully unnested director's name. but if we check the shape of the dataframe it will increase because after unnesting it
df_director.shape

Out[9]: (9612, 3)

In [10]: # Now we will do unnesting for cast column

df_cast = df[["title", "cast"]]
df_cast["cast_unnested"] = df_cast["cast"].apply(lambda a: str(a).split(", "))
df_cast = df_cast.explode("cast_unnested")
df_cast.head(5)

Out[10]: title cast cast_unnested

0 Dick Johnson Is Dead NaN nan

1 Blood & Water Ama Qamata, Khosi Ngema, Gail Mabalane, Thaban... Ama Qamata

1 Blood & Water Ama Qamata, Khosi Ngema, Gail Mabalane, Thaban... Khosi Ngema

1 Blood & Water Ama Qamata, Khosi Ngema, Gail Mabalane, Thaban... Gail Mabalane

1 Blood & Water Ama Qamata, Khosi Ngema, Gail Mabalane, Thaban... Thabang Molaba

In [11]: # Now we will do unnesting for country column

df_country = df[["title", "country"]]
df_country["country_unnested"] = df_country["country"].apply(lambda a: str(a).split(", "))
df_country = df_country.explode("country_unnested")
df_country.head(5)

3 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[11]: title country country_unnested

0 Dick Johnson Is Dead United States United States

1 Blood & Water South Africa South Africa

2 Ganglands NaN nan

3 Jailbirds New Orleans NaN nan

4 Kota Factory India India

In [12]: # Now we will do unnesting for listed_in column

df_listed_in = df[["title", "listed_in"]]
df_listed_in["listed_in_unnested"] = df_listed_in["listed_in"].apply(lambda a: str(a).split(", "))
df_listed_in = df_listed_in.explode("listed_in_unnested")
df_listed_in.head(5)

Out[12]: title listed_in listed_in_unnested

0 Dick Johnson Is Dead Documentaries Documentaries

1 Blood & Water International TV Shows, TV Dramas, TV Mysteries International TV Shows

1 Blood & Water International TV Shows, TV Dramas, TV Mysteries TV Dramas

1 Blood & Water International TV Shows, TV Dramas, TV Mysteries TV Mysteries

2 Ganglands Crime TV Shows, International TV Shows, TV Act... Crime TV Shows

In [13]: # we have unnested all 4 columns. now we have to merge the unnested values with the original dataframe
# first we are going to join the original dataframe with df_director
# then we drop the duplicate columns director_x, director_y
# then we rename the column director_unnested to director
df_final = df.merge(df_director, on= "title", how = "left")
df_final.drop(["director_x","director_y"], axis = 1, inplace = True)
df_final.rename(columns = {"director_unnested": "director"}, inplace = True)
df_final.head(5)

Out[13]: show_id type title cast country date_added release_year rating duration listed_in description director

As her
Dick father nears
United September Kirsten
0 s1 Movie Johnson Is NaN 2020 PG-13 90 min Documentaries the end of
States 25, 2021 Johnson
Dead his life,
filmm...

Ama
After
Qamata,
International crossing
Khosi
TV Blood & South September TV- 2 TV Shows, TV paths at a
1 s2 Ngema, 2021 nan
Show Water Africa 24, 2021 MA Seasons Dramas, TV party, a
Gail
Mysteries Cape Town
Mabalane,
t...
Thaban...

Sami
Bouajila, Crime TV To protect
Tracy Shows, his family
TV September TV- Julien
2 s3 Ganglands Gotoas, NaN 2021 1 Season International from a
Show 24, 2021 MA Leclercq
Samuel TV Shows, TV powerful
Jouy, Act... drug lor...
Nabi...

Feuds,
Jailbirds flirtations
TV September TV- Docuseries,
3 s4 New NaN NaN 2021 1 Season and toilet nan
Show 24, 2021 MA Reality TV
Orleans talk go
down amo...

Mayur
In a city of
More, International
coaching
TV Kota Jitendra September TV- 2 TV Shows,
4 s5 India 2021 centers nan
Show Factory Kumar, 24, 2021 MA Seasons Romantic TV
known to
Ranjan Raj, Shows, TV ...
train I...
Alam K...

In [14]: df_final.shape

Out[14]: (9612, 12)

In [15]: # Now we are going to join the original dataframe with df_cast
# Then we drop the duplicate columns cast_x, cast_y
# Then we rename the column cast_unnested to cast
df_final = df_final.merge(df_cast, on= "title", how = "left")
df_final.drop(["cast_x","cast_y"], axis = 1, inplace = True)
df_final.rename(columns = {"cast_unnested": "cast"}, inplace = True)

4 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

df_final.head(5)

Out[15]: show_id type title country date_added release_year rating duration listed_in description director cast

As her father
Dick
United September nears the end Kirsten
0 s1 Movie Johnson 2020 PG-13 90 min Documentaries nan
States 25, 2021 of his life, Johnson
Is Dead
filmm...

International TV After crossing

TV Blood & South September TV- 2 Shows, TV paths at a Ama
1 s2 2021 nan
Show Water Africa 24, 2021 MA Seasons Dramas, TV party, a Cape Qamata
Mysteries Town t...

International TV After crossing

TV Blood & South September TV- 2 Shows, TV paths at a Khosi
2 s2 2021 nan
Show Water Africa 24, 2021 MA Seasons Dramas, TV party, a Cape Ngema
Mysteries Town t...

International TV After crossing

TV Blood & South September TV- 2 Shows, TV paths at a Gail
3 s2 2021 nan
Show Water Africa 24, 2021 MA Seasons Dramas, TV party, a Cape Mabalane
Mysteries Town t...

International TV After crossing

TV Blood & South September TV- 2 Shows, TV paths at a Thabang
4 s2 2021 nan
Show Water Africa 24, 2021 MA Seasons Dramas, TV party, a Cape Molaba
Mysteries Town t...

In [16]: df_final.shape

Out[16]: (70812, 12)

In [17]: # Now we are going to join the original dataframe with df_country
# Then we drop the duplicate columns country_x, country_y
# Then we rename the column country_unnested to country
df_final = df_final.merge(df_country, on= "title", how = "left")
df_final.drop(["country_x","country_y"], axis = 1, inplace = True)
df_final.rename(columns = {"country_unnested": "country"}, inplace = True)
df_final.head(5)

Out[17]: show_id type title date_added release_year rating duration listed_in description director cast country

As her father
Dick
September nears the end Kirsten United
0 s1 Movie Johnson 2020 PG-13 90 min Documentaries nan
25, 2021 of his life, Johnson States
Is Dead
filmm...

International TV After crossing

TV Blood & September TV- 2 Shows, TV paths at a Ama South
1 s2 2021 nan
Show Water 24, 2021 MA Seasons Dramas, TV party, a Cape Qamata Africa
Mysteries Town t...

International TV After crossing

TV Blood & September TV- 2 Shows, TV paths at a Khosi South
2 s2 2021 nan
Show Water 24, 2021 MA Seasons Dramas, TV party, a Cape Ngema Africa
Mysteries Town t...

International TV After crossing

TV Blood & September TV- 2 Shows, TV paths at a Gail South
3 s2 2021 nan
Show Water 24, 2021 MA Seasons Dramas, TV party, a Cape Mabalane Africa
Mysteries Town t...

International TV After crossing

TV Blood & September TV- 2 Shows, TV paths at a Thabang South
4 s2 2021 nan
Show Water 24, 2021 MA Seasons Dramas, TV party, a Cape Molaba Africa
Mysteries Town t...

In [18]: df_final.shape

Out[18]: (89382, 12)

In [19]: # Now we are going to join the original dataframe with df_listed_in
# Then we drop the duplicate columns listed_in_x, listed_in_y
# Then we rename the column listed_in_unnested to listed_in
df_final = df_final.merge(df_listed_in, on= "title", how = "left")
df_final.drop(["listed_in_x","listed_in_y"], axis = 1, inplace = True)
df_final.rename(columns = {"listed_in_unnested": "listed_in"}, inplace = True)
df_final.head(5)

5 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[19]: show_id type title date_added release_year rating duration description director cast country listed_in

As her father
Dick
September nears the end Kirsten United
0 s1 Movie Johnson 2020 PG-13 90 min nan Documentaries
25, 2021 of his life, Johnson States
Is Dead
filmm...

After crossing
TV Blood & September TV- 2 paths at a Ama South International
1 s2 2021 nan
Show Water 24, 2021 MA Seasons party, a Cape Qamata Africa TV Shows
Town t...

After crossing
TV Blood & September TV- 2 paths at a Ama South
2 s2 2021 nan TV Dramas
Show Water 24, 2021 MA Seasons party, a Cape Qamata Africa
Town t...

After crossing
TV Blood & September TV- 2 paths at a Ama South
3 s2 2021 nan TV Mysteries
Show Water 24, 2021 MA Seasons party, a Cape Qamata Africa
Town t...

After crossing
TV Blood & September TV- 2 paths at a Khosi South International
4 s2 2021 nan
Show Water 24, 2021 MA Seasons party, a Cape Ngema Africa TV Shows
Town t...

In [20]: # Finally after unnesting and merging the 4 columns with the original dataframe now we see we have 327001 rows.
df_final.shape

Out[20]: (201991, 12)

UNIQUE ATTRIBUTES
In [21]: #unique attributes in Director Column:
print("Unique Directors: ", df_final["director"].unique())
print("Total No of Directors: ", df_final["director"].nunique())

Unique Directors: ['Kirsten Johnson' 'nan' 'Julien Leclercq' ... 'Majid Al Ansari'
'Peter Hewitt' 'Mozez Singh']
Total No of Directors: 4994

In [22]: #unique attributes in Cast Column:

print("Unique Actors: ", df_final["cast"].unique())
print("Total No of Actors: ", df_final["cast"].nunique())

Unique Actors: ['nan' 'Ama Qamata' 'Khosi Ngema' ... 'Malkeet Rauni' 'Anita Shabdish'
'Chittaranjan Tripathy']
Total No of Actors: 36440

In [23]: #unique attributes in Listed_in Column:

print("Unique Genre: ", df_final["listed_in"].unique())
print("Total No of Genre: ", df_final["listed_in"].nunique())

Unique Genre: ['Documentaries' 'International TV Shows' 'TV Dramas' 'TV Mysteries'

'Crime TV Shows' 'TV Action & Adventure' 'Docuseries' 'Reality TV'
'Romantic TV Shows' 'TV Comedies' 'TV Horror' 'Children & Family Movies'
'Dramas' 'Independent Movies' 'International Movies' 'British TV Shows'
'Comedies' 'Spanish-Language TV Shows' 'Thrillers' 'Romantic Movies'
'Music & Musicals' 'Horror Movies' 'Sci-Fi & Fantasy' 'TV Thrillers'
"Kids' TV" 'Action & Adventure' 'TV Sci-Fi & Fantasy' 'Classic Movies'
'Anime Features' 'Sports Movies' 'Anime Series' 'Korean TV Shows'
'Science & Nature TV' 'Teen TV Shows' 'Cult Movies' 'TV Shows'
'Faith & Spirituality' 'LGBTQ Movies' 'Stand-Up Comedy' 'Movies'
'Stand-Up Comedy & Talk Shows' 'Classic & Cult TV']
Total No of Genre: 42

In [24]: #unique attributes in Rating Column:

print("Unique Ratings: ", df_final["rating"].unique())
print("Total No of Ratings: ", df_final["rating"].nunique())

Unique Ratings: ['PG-13' 'TV-MA' 'PG' 'TV-14' 'TV-PG' 'TV-Y' 'TV-Y7' 'R' 'TV-G' 'G'
'NC-17' '74 min' '84 min' '66 min' 'NR' nan 'TV-Y7-FV' 'UR']
Total No of Ratings: 17

In [25]: #unique attributes in Country Column:

print("Unique Country: ", df_final["country"].unique())
print("Total No of Country: ", df_final["country"].nunique())

6 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Unique Country: ['United States' 'South Africa' 'nan' 'India' 'Ghana' 'Burkina Faso'
'United Kingdom' 'Germany' 'Ethiopia' 'Czech Republic' 'Mexico' 'Turkey'
'Australia' 'France' 'Finland' 'China' 'Canada' 'Japan' 'Nigeria' 'Spain'
'Belgium' 'South Korea' 'Singapore' 'Italy' 'Romania' 'Argentina'
'Venezuela' 'Hong Kong' 'Russia' '' 'Ireland' 'Nepal' 'New Zealand'
'Brazil' 'Greece' 'Jordan' 'Colombia' 'Switzerland' 'Israel' 'Taiwan'
'Bulgaria' 'Algeria' 'Poland' 'Saudi Arabia' 'Thailand' 'Indonesia'
'Egypt' 'Denmark' 'Kuwait' 'Netherlands' 'Malaysia' 'Vietnam' 'Hungary'
'Sweden' 'Lebanon' 'Syria' 'Philippines' 'Iceland' 'United Arab Emirates'
'Norway' 'Qatar' 'Mauritius' 'Austria' 'Cameroon' 'Palestine' 'Uruguay'
'United Kingdom,' 'Kenya' 'Chile' 'Luxembourg' 'Cambodia' 'Bangladesh'
'Portugal' 'Cayman Islands' 'Senegal' 'Serbia' 'Malta' 'Namibia' 'Angola'
'Peru' 'Mozambique' 'Cambodia,' 'Belarus' 'Zimbabwe' 'Puerto Rico'
'Pakistan' 'Cyprus' 'Guatemala' 'Iraq' 'Malawi' 'Paraguay' 'Croatia'
'Iran' 'West Germany' 'United States,' 'Albania' 'Georgia' 'Soviet Union'
'Morocco' 'Slovakia' 'Ukraine' 'Bermuda' 'Ecuador' 'Armenia' 'Mongolia'
'Bahamas' 'Sri Lanka' 'Latvia' 'Liechtenstein' 'Cuba' 'Nicaragua'
'Poland,' 'Slovenia' 'Dominican Republic' 'Samoa' 'Azerbaijan' 'Botswana'
'Vatican City' 'Jamaica' 'Kazakhstan' 'Lithuania' 'Afghanistan' 'Somalia'
'Sudan' 'Panama' 'Uganda' 'East Germany' 'Montenegro']
Total No of Country: 128

In [26]: # There are 36440 Actors in the data Frame

df_final["cast"].nunique()

Out[26]: 36440

In [27]: # To Avoid the missing Values we replace Nan with Unknown

df_final['cast'].replace(["nan"], ["Unknown Actor"], inplace = True)
df_final['director'].replace(["nan"], ["Unknown Directors"], inplace = True)
df_final['country'].replace(["nan"], [np.nan], inplace =True)

In [28]: df_final.head(5)

Out[28]: show_id type title date_added release_year rating duration description director cast country listed_in

As her father
Dick
September nears the end Kirsten Unknown United
0 s1 Movie Johnson 2020 PG-13 90 min Documentaries
25, 2021 of his life, Johnson Actor States
Is Dead
filmm...

After crossing
TV Blood & September TV- 2 paths at a Unknown Ama South International
1 s2 2021
Show Water 24, 2021 MA Seasons party, a Cape Directors Qamata Africa TV Shows
Town t...

After crossing
TV Blood & September TV- 2 paths at a Unknown Ama South
2 s2 2021 TV Dramas
Show Water 24, 2021 MA Seasons party, a Cape Directors Qamata Africa
Town t...

After crossing
TV Blood & September TV- 2 paths at a Unknown Ama South
3 s2 2021 TV Mysteries
Show Water 24, 2021 MA Seasons party, a Cape Directors Qamata Africa
Town t...

After crossing
TV Blood & September TV- 2 paths at a Unknown Khosi South International
4 s2 2021
Show Water 24, 2021 MA Seasons party, a Cape Directors Ngema Africa TV Shows
Town t...

REMOVING/FILLING NULL OR MISSING VALUES

In [29]: df_final.isnull().sum()

Out[29]: show_id 0
type 0
title 0
date_added 158
release_year 0
rating 67
duration 3
description 0
director 0
cast 0
country 11897
listed_in 0
dtype: int64

In [30]: #we found that there are three rows where the duration is typed in rating Column
df_final[df_final['duration'].isnull()]

7 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[30]: show_id type title date_added release_year rating duration description director cast country listed_in

Louis C.K. muses

Louis C.K. 74 Louis Louis United
126537 s5542 Movie April 4, 2017 2017 NaN on religion, Movies
2017 min C.K. C.K. States
eternal love, gi...

Emmy-winning
Louis C.K.: September 84 comedy writer Louis Louis United
131603 s5795 Movie 2010 NaN Movies
Hilarious 16, 2016 min Louis C.K. brings C.K. C.K. States
h...

Louis C.K.: The comic puts

Live at the August 15, 66 his trademark Louis Louis United
131737 s5814 Movie 2015 NaN Movies
Comedy 2016 min hilarious/ C.K. C.K. States
Store thought...

In [31]: # we now changing the mistaken rating column to the duration

df_final.loc[df_final['duration'].isnull(), 'duration'] = df_final.loc[df_final['duration'].isnull(), 'duration'].fillna(df_final

In [32]: df_final[df_final['duration'].isnull()]

Out[32]: show_id type title date_added release_year rating duration description director cast country listed_in

In [33]: # now we are filling the missing values of rating Column

df_final.loc[df_final['rating'].str.contains('min', na =False),'rating'] = 'NR'
df_final['rating'].fillna('NR', inplace = True)
df_final.head()

Out[33]: show_id type title date_added release_year rating duration description director cast country listed_in

As her father
Dick
September nears the end Kirsten Unknown United
0 s1 Movie Johnson 2020 PG-13 90 min Documentaries
25, 2021 of his life, Johnson Actor States
Is Dead
filmm...

After crossing
TV Blood & September TV- 2 paths at a Unknown Ama South International
1 s2 2021
Show Water 24, 2021 MA Seasons party, a Cape Directors Qamata Africa TV Shows
Town t...

After crossing
TV Blood & September TV- 2 paths at a Unknown Ama South
2 s2 2021 TV Dramas
Show Water 24, 2021 MA Seasons party, a Cape Directors Qamata Africa
Town t...

After crossing
TV Blood & September TV- 2 paths at a Unknown Ama South
3 s2 2021 TV Mysteries
Show Water 24, 2021 MA Seasons party, a Cape Directors Qamata Africa
Town t...

After crossing
TV Blood & September TV- 2 paths at a Unknown Khosi South International
4 s2 2021
Show Water 24, 2021 MA Seasons party, a Cape Directors Ngema Africa TV Shows
Town t...

In [34]: df_final.isna().sum()

Out[34]: show_id 0
type 0
title 0
date_added 158
release_year 0
rating 0
duration 0
description 0
director 0
cast 0
country 11897
listed_in 0
dtype: int64

8 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

In [35]: df_final['date_added']

Out[35]: 0 September 25, 2021

1 September 24, 2021
2 September 24, 2021
3 September 24, 2021
4 September 24, 2021
...
201986 March 2, 2019
201987 March 2, 2019
201988 March 2, 2019
201989 March 2, 2019
201990 March 2, 2019
Name: date_added, Length: 201991, dtype: object

DATA TYPE CONVERSION (OBJECT TYPE TO DATETIME TYPE)

In [36]: # Date added column is in object data type. we need that in datetime date type
df_final['date_added']=pd.to_datetime(df_final['date_added'],format='mixed')

In [37]: df_final['date_added']

Out[37]: 0 2021-09-25
1 2021-09-24
2 2021-09-24
3 2021-09-24
4 2021-09-24
...
201986 2019-03-02
201987 2019-03-02
201988 2019-03-02
201989 2019-03-02
201990 2019-03-02
Name: date_added, Length: 201991, dtype: datetime64[ns]

In [38]: df_final.isna().sum()

Out[38]: show_id 0
type 0
title 0
date_added 158
release_year 0
rating 0
duration 0
description 0
director 0
cast 0
country 11897
listed_in 0
dtype: int64

In [39]: # we are filling the missing country rows. when the same director have country name in some other entry we are imputing them to the miss
for i in df_final[df_final['country'].isnull()]['director'].unique():
if i in df_final[~df_final['country'].isnull()]['director'].unique():
imp=df_final[df_final['director']==i]['country'].mode().values[0]
df_final.loc[df_final['director']==i,'country']=df_final.loc[df_final['director']==i,'country'].fillna(imp)

In [40]: # we are filling the missing country rows . when the same cast have country name in some other entry we are imputing them to the missing
for i in df_final[df_final['country'].isnull()]['cast'].unique():
if i in df_final[~df_final['country'].isnull()]['cast'].unique():
imp=df_final[df_final['cast']==i]['country'].mode().values[0]
df_final.loc[df_final['cast']==i,'country']=df_final.loc[df_final['cast']==i,'country'].fillna(imp)

In [41]: #after imputation also there are some missing values so we fill unknown country for the remaining values
df_final["country"].fillna("Unknown Country",inplace=True)

In [42]: df_final.isna().sum()

Out[42]: show_id 0
type 0
title 0
date_added 158
release_year 0
rating 0
duration 0
description 0
director 0
cast 0
country 0
listed_in 0
dtype: int64

In [43]: # we are filling the missing date_added rows.

for i in df_final[df_final['date_added'].isnull()]['release_year'].unique():

9 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

imp=df_final[df_final['release_year']==i]['date_added'].mode().values[0]
df_final.loc[df_final['release_year']==i,'date_added']=df_final.loc[df_final['release_year']==i,'date_added'].fillna(imp)

In [44]: #now data cleaning is done successfully. we can see there are no missing values
df_final.isna().sum()

Out[44]: show_id 0
type 0
title 0
date_added 0
release_year 0
rating 0
duration 0
description 0
director 0
cast 0
country 0
listed_in 0
dtype: int64

In [45]: df_final["duration"].value_counts()

Out[45]: duration
1 Season 35035
2 Seasons 9559
3 Seasons 5084
94 min 4343
106 min 4040
...
3 min 4
5 min 3
11 min 2
8 min 2
9 min 2
Name: count, Length: 220, dtype: int64

In [46]: # we cannot use duration column for EDA when it has Season/ Minutes with the data. so we replace the Season/ Minutes keyword
df_final["duration"] = df_final["duration"].str.replace("min", "")
df_final["duration"] = df_final["duration"].str.replace("Seasons", "")
df_final["duration"] = df_final["duration"].str.replace("Season", "")
df_final["duration"] = df_final["duration"].astype("int")
df_final.head()

Out[46]: show_id type title date_added release_year rating duration description director cast country listed_in

As her father
Dick
nears the end Kirsten Unknown United
0 s1 Movie Johnson 2021-09-25 2020 PG-13 90 Documentaries
of his life, Johnson Actor States
Is Dead
filmm...

After crossing
TV Blood & TV- paths at a Unknown Ama South International
1 s2 2021-09-24 2021 2
Show Water MA party, a Cape Directors Qamata Africa TV Shows
Town t...

After crossing
TV Blood & TV- paths at a Unknown Ama South
2 s2 2021-09-24 2021 2 TV Dramas
Show Water MA party, a Cape Directors Qamata Africa
Town t...

After crossing
TV Blood & TV- paths at a Unknown Ama South
3 s2 2021-09-24 2021 2 TV Mysteries
Show Water MA party, a Cape Directors Qamata Africa
Town t...

After crossing
TV Blood & TV- paths at a Unknown Khosi South International
4 s2 2021-09-24 2021 2
Show Water MA party, a Cape Directors Ngema Africa TV Shows
Town t...

In [47]: df_final["duration"].unique()

10 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[47]: array([ 90, 2, 1, 91, 125, 9, 104, 127, 4, 67, 94, 5, 161,
61, 166, 147, 103, 97, 106, 111, 3, 110, 105, 96, 124, 116,
98, 23, 115, 122, 99, 88, 100, 6, 102, 93, 95, 85, 83,
113, 13, 182, 48, 145, 87, 92, 80, 117, 128, 119, 143, 114,
118, 108, 63, 121, 142, 154, 120, 82, 109, 101, 86, 229, 76,
89, 156, 112, 107, 129, 135, 136, 165, 150, 133, 70, 84, 140,
78, 7, 64, 59, 139, 69, 148, 189, 141, 130, 138, 81, 132,
10, 123, 65, 68, 66, 62, 74, 131, 39, 46, 38, 8, 17,
126, 155, 159, 137, 12, 273, 36, 34, 77, 60, 49, 58, 72,
204, 212, 25, 73, 29, 47, 32, 35, 71, 149, 33, 15, 54,
224, 162, 37, 75, 79, 55, 158, 164, 173, 181, 185, 21, 24,
51, 151, 42, 22, 134, 177, 52, 14, 53, 57, 28, 50, 26,
45, 171, 27, 44, 146, 20, 157, 203, 41, 30, 194, 233, 237,
230, 195, 253, 152, 190, 160, 208, 180, 144, 174, 170, 192, 209,
187, 172, 16, 186, 11, 193, 176, 56, 169, 40, 168, 312, 153,
214, 31, 163, 19, 179, 43, 200, 196, 167, 178, 228, 18, 205,
201, 191])

Value_counts Of Each Columns

In [54]: #Value Counts For Movies in year Wise
df_value_release_year_m = df_movies.groupby("release_year").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False
df_value_release_year_m.head()

Out[54]: release_year title

0 2018 767

1 2017 767

2 2016 658

3 2019 633

4 2020 517

In [55]: #Value Counts For TV Shows in year Wise

df_value_release_year_t = df_tv.groupby("release_year").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).
df_value_release_year_t.head()

Out[55]: release_year title

0 2020 436

1 2019 397

2 2018 380

3 2021 315

4 2017 265

In [56]: #Value Counts For Movies in Rating Wise

df_value_rating_m = df_movies.groupby("rating").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_value_rating_m.head(5)

Out[56]: rating title

0 TV-MA 2062

1 TV-14 1427

2 R 797

3 TV-PG 540

4 PG-13 490

In [57]: #Value Counts For TV Shows in Rating Wise

df_value_rating_t = df_tv.groupby("rating").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
df_value_rating_t.head(5)

Out[57]: rating title

0 TV-MA 1145

1 TV-14 733

2 TV-PG 323

3 TV-Y7 195

4 TV-Y 176

11 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

In [58]: #Value Counts For Movies in Duration Wise

df_value_duration_m = df_movies.groupby("duration").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_value_duration_m.head(5)

Out[58]: duration title

0 90 152

1 97 146

2 93 146

3 94 146

4 91 144

In [59]: #Value Counts For TV Shows in Duration Wise

df_value_duration_t = df_tv.groupby("duration").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_value_duration_t.head(5)

Out[59]: duration title

0 1 1793

1 2 425

2 3 199

3 4 95

4 5 65

In [60]: #Value Counts For Movies in Country Wise

df_value_country_m = df_movies.groupby("country").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_value_country_m = df_value_country_m[df_value_country_m["country"]!= "Unknown Country"]
df_value_country_m.head(5)

Out[60]: country title

0 United States 2940

1 India 1052

2 United Kingdom 556

3 Canada 334

4 France 318

In [61]: #Value Counts For TV Shows in Country Wise

df_value_country_t = df_tv.groupby("country").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_value_country_t = df_value_country_t[df_value_country_t["country"]!= "Unknown Country"]
df_value_country_t.head(5)

Out[61]: country title

0 United States 1308

1 United Kingdom 273

2 Japan 200

3 South Korea 171

4 Canada 126

In [62]: #Value Counts For Movies in Listed_in Wise

df_value_list_m = df_movies.groupby("listed_in").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_value_list_m.head(5)

Out[62]: listed_in title

0 International Movies 2752

1 Dramas 2427

2 Comedies 1674

3 Documentaries 869

4 Action & Adventure 859

In [63]: #Value Counts For TV Shows in Listed_in Wise

df_value_list_t = df_tv.groupby("listed_in").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_value_list_t.head(5)

12 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[63]: listed_in title

0 International TV Shows 1351

1 TV Dramas 763

2 TV Comedies 581

3 Crime TV Shows 470

4 Kids' TV 451

Statistical Summary
In [64]: # Statistical Summary of Duration Column in Movies
df_value_duration_m["duration"].describe()

Out[64]: count 205.000000

mean 110.395122
std 63.942797
min 3.000000
25% 57.000000
50% 108.000000
75% 159.000000
max 312.000000
Name: duration, dtype: float64

we can see the minimum duration of a movie is 3 minutes and maximum Duration is 312 minutes

In [65]: # Statistical Summary of Duration Column in TV Shows

df_value_duration_t["duration"].describe()

Out[65]: count 15.000000

mean 8.200000
std 4.813671
min 1.000000
25% 4.500000
50% 8.000000
75% 11.500000
max 17.000000
Name: duration, dtype: float64

we can see the minimum Season for a TV Show is 1 Season and maximum is 17 Season

In [66]: #we are splitting the date_added column into mutiple columns for EDA purpose
df_final['year'] = df_final['date_added'].dt.year
df_final['month'] = df_final['date_added'].dt.month
df_final['day'] = df_final['date_added'].dt.day
df_final['day_name'] = df_final['date_added'].dt.day_name()

In [67]: df_final.head()

13 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[67]: show_id type title date_added release_year rating duration description director cast country listed_in year month

As her
Dick father nears
Kirsten Unknown United
0 s1 Movie Johnson 2021-09-25 2020 PG-13 90 the end of Documentaries 2021
Johnson Actor States
Is Dead his life,
filmm...

After
crossing
Blood
TV TV- paths at a Unknown Ama South International
1 s2 & 2021-09-24 2021 2 2021
Show MA party, a Directors Qamata Africa TV Shows
Water
Cape Town
t...

After
crossing
Blood
TV TV- paths at a Unknown Ama South
2 s2 & 2021-09-24 2021 2 TV Dramas 2021
Show MA party, a Directors Qamata Africa
Water
Cape Town
t...

After
crossing
Blood
TV TV- paths at a Unknown Ama South
3 s2 & 2021-09-24 2021 2 TV Mysteries 2021
Show MA party, a Directors Qamata Africa
Water
Cape Town
t...

After
crossing
Blood
TV TV- paths at a Unknown Khosi South International
4 s2 & 2021-09-24 2021 2 2021
Show MA party, a Directors Ngema Africa TV Shows
Water
Cape Town
t...

In [68]: # Seperated the movies from the final dataframe

df_movies = df_final[df_final["type"]== "Movie"]
df_movies.head(5)

Out[68]: show_id type title date_added release_year rating duration description director cast country listed_in year

As her
Dick father nears
Kirsten Unknown United
0 s1 Movie Johnson Is 2021-09-25 2020 PG-13 90 the end of Documentaries 2021
Johnson Actor States
Dead his life,
filmm...

Equestria's
My Little
divided. But
Pony: A Robert Vanessa United Children &
159 s7 Movie 2021-09-24 2021 PG 91 a bright- 2021
New Cullen Hudgens States Family Movies
eyed hero
Generation
be...

Equestria's
My Little
divided. But
Pony: A Robert Kimiko United Children &
160 s7 Movie 2021-09-24 2021 PG 91 a bright- 2021
New Cullen Glenn States Family Movies
eyed hero
Generation
be...

Equestria's
My Little
divided. But
Pony: A Robert James United Children &
161 s7 Movie 2021-09-24 2021 PG 91 a bright- 2021
New Cullen Marsden States Family Movies
eyed hero
Generation
be...

Equestria's
My Little
divided. But
Pony: A Robert Sofia United Children &
162 s7 Movie 2021-09-24 2021 PG 91 a bright- 2021
New Cullen Carson States Family Movies
eyed hero
Generation
be...

In [69]: # Starting Year and Ending Year of Movies:

print("Minimum Year : ",df_movies["release_year"].min())
print("Maximum Year : ",df_movies["release_year"].max())

Minimum Year : 1942

Maximum Year : 2021

In [70]: # Starting Year and Ending Year of TV Shows:

print("Minimum Year : ",df_tv["release_year"].min())
print("Maximum Year : ",df_tv["release_year"].max())

Minimum Year : 1925

Maximum Year : 2021

14 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

In [71]: # Seperated the tv shows from the final dataframe

df_tv = df_final[df_final["type"]== "TV Show"]
df_tv.head(5)

Out[71]: show_id type title date_added release_year rating duration description director cast country listed_in year month

After
crossing
Blood
TV TV- paths at a Unknown Ama South International
1 s2 & 2021-09-24 2021 2 2021 9
Show MA party, a Directors Qamata Africa TV Shows
Water
Cape Town
t...

After
crossing
Blood
TV TV- paths at a Unknown Ama South
2 s2 & 2021-09-24 2021 2 TV Dramas 2021 9
Show MA party, a Directors Qamata Africa
Water
Cape Town
t...

After
crossing
Blood
TV TV- paths at a Unknown Ama South TV
3 s2 & 2021-09-24 2021 2 2021 9
Show MA party, a Directors Qamata Africa Mysteries
Water
Cape Town
t...

After
crossing
Blood
TV TV- paths at a Unknown Khosi South International
4 s2 & 2021-09-24 2021 2 2021 9
Show MA party, a Directors Ngema Africa TV Shows
Water
Cape Town
t...

After
crossing
Blood
TV TV- paths at a Unknown Khosi South
5 s2 & 2021-09-24 2021 2 TV Dramas 2021 9
Show MA party, a Directors Ngema Africa
Water
Cape Town
t...

Now we are going to do Exploratory Data Analysis for the Netflix

Dataset:
I) Uni Variate Analysis for the follwing variables:

1) Top 5 Genre in TV Shows & Movies

2) Top 5 Directors in TV Shows & Movies
3) Top 5 Cast in TV Shows & Movies
4) Highest No.of TV Shows & Movies Added in Netflix in Years
5) No.of TV Shows & Movies Released in Years
6) Top 5 Ratings in TV Shows & Movies
7) Top 5 Actors in Indian TV Shows & Movies
8) Top 5 Directors in Indian TV Shows & Movies
9) TOP 10 Countries Releasing TV SHOWS & MOVIES

TOP 5 Genre in TV SHOWS & MOVIES

In [73]: #TOP 5 Genre in TV SHOWS & MOVIES
#Graphical RepresentatioTn using bar plot
fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
df_genre = df_tv.groupby("listed_in").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index().head
x = df_genre["listed_in"]
y = df_genre["title"]
height = plt.bar(x,y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.xticks(rotation = 40)
plt.title("Top 5 Genre in TV Shows")
plt.xlabel("Top Genre in Tv Shows")
plt.ylabel("Count")

plt.subplot (2,2,2)
df_genre_m = df_movies.groupby("listed_in").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
x = df_genre_m["listed_in"]
y = df_genre_m["title"]
height = plt.bar(x,y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()

15 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")

plt.xticks(rotation = 40)
plt.title("Top 5 Genre in MOVIES")
plt.xlabel("Top Genre in Movies")
plt.ylabel("Count")
plt.suptitle("TOP 5 Genre in TV SHOWS & MOVIES", fontsize = 24)
plt.show()

TOP 5 Directors in TV SHOWS & MOVIES

In [74]: # to find which director had more movies
df_more_movies = df_movies.groupby("director").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_more_movies = df_more_movies[df_more_movies["director"]!= "Unknown Directors"]
df_more_movies.head(5)

Out[74]: director title

1 Rajiv Chilaka 22

2 Jan Suter 21

3 Raúl Campos 19

4 Suhas Kadav 16

5 Marcus Raboy 15

In [75]: # to find which director had more TV Shows

df_more_tv = df_tv.groupby("director").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
df_more_tv = df_more_tv[df_more_tv["director"]!= "Unknown Directors"]
df_more_tv.head(5)

Out[75]: director title

1 Ken Burns 3

2 Alastair Fothergill 3

3 Stan Lathan 2

4 Joe Berlinger 2

5 Hsu Fu-chun 2

In [76]: #TOP 5 Directors in TV SHOWS & MOVIES

#Graphical RepresentatioTn using bar plot
fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
df_more_tv_top5 = df_more_tv.head(5)
x= df_more_tv_top5["director"]
y = df_more_tv_top5["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Directors Having MORE TV SHOWS")
plt.xlabel("Top Directors")
plt.ylabel("No.of TV Shows")

16 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

plt.xticks(rotation = 15)

plt.subplot (2,2,2)
df_more_movies_top5 = df_more_movies.head(5)
x= df_more_movies_top5["director"]
y = df_more_movies_top5["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Directors Having MORE MOVIES")
plt.xlabel("Top Directors")
plt.ylabel("No.of Movies")
plt.xticks(rotation = 15)
plt.suptitle("TOP 5 Directors in TV SHOWS & MOVIES", fontsize = 24)
plt.show()

TOP 5 Cast in TV SHOWS & MOVIES

In [77]: # to find Top 5 Cast in Movies
df_c_movies = df_movies.groupby("cast").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
df_c_movies = df_c_movies[df_c_movies["cast"]!= "Unknown Actor"]
df_c_movies.head(5)

Out[77]: cast title

1 Anupam Kher 42

2 Shah Rukh Khan 35

3 Naseeruddin Shah 32

4 Om Puri 30

5 Akshay Kumar 30

In [78]: # to find Top 5 Cast in TV Shows

df_c_tv = df_tv.groupby("cast").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
df_c_tv = df_c_tv[df_c_tv["cast"]!= "Unknown Actor"]
df_c_tv.head(5)

Out[78]: cast title

1 Takahiro Sakurai 25

2 Yuki Kaji 19

3 Junichi Suwabe 17

4 Daisuke Ono 17

5 Ai Kayano 17

In [79]: #TOP 5 Cast in TV SHOWS & MOVIES

#Graphical RepresentatioTn using bar plot
fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
df_c_tv_top5 = df_c_tv.head(5)
x= df_c_tv_top5["cast"]
y = df_c_tv_top5["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))

17 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

for bar in height:

yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Cast in TV SHOWS")
plt.xlabel("Top Cast")
plt.ylabel("No.of TV Shows")
plt.xticks(rotation = 15)

plt.subplot (2,2,2)
df_c_movies_top5 = df_c_movies.head(5)
x= df_c_movies_top5["cast"]
y = df_c_movies_top5["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Cast in MOVIES")
plt.xlabel("Top Cast")
plt.ylabel("No.of Movies")
plt.xticks(rotation = 15)
plt.suptitle("TOP 5 Cast in TV SHOWS & MOVIES", fontsize = 24)
plt.show()

Highest No.Of Shows added in Netflix in the Year 2020

In [80]: # In which year highest no.of TV Schows were Added?
df_tv_year = df_tv.groupby("year").agg({"title":"nunique"}).reset_index()
df_tv_year.head()

Out[80]: year title

0 2008 1

1 2013 5

2 2014 5

3 2015 26

4 2016 179

Highest No.Of Movies was added in Netflix in the Year 2019

In [81]: # In which year highest no.of Movies were Added?
df_mov_year = df_movies.groupby("year").agg({"title":"nunique"}).reset_index()
df_mov_year.head()

Out[81]: year title

0 2008 1

1 2009 2

2 2010 1

3 2011 13

4 2012 3

18 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

In [82]: #Graphical Representation using bar plot

fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
x = df_tv_year["year"]
y = df_tv_year["title"]
height = plt.bar(x,y, color ="darkorange")
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("No.Of TV Shows Added in Years")
plt.xlabel("Year")
plt.ylabel("Count Of Shows")

plt.subplot (2,2,2)
x = df_mov_year["year"]
y = df_mov_year["title"]
height = plt.bar(x,y, color ="royalblue")
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("No.Of Movies Added in Years")
plt.xlabel("Year")
plt.ylabel("Count Of Movies")
plt.suptitle("No Of TV SHOWS & MOVIES Added in Netflix ", fontsize = 24)
plt.show()

Highest No.Of Movies released in the Year 2017 & 2018

In [83]: # In which year highest no.of Movies were released?
df_mov_y = df_movies.groupby("release_year").agg({"title":"nunique"}).reset_index()
df_mov_y.head()

Out[83]: release_year title

0 1942 2

1 1943 3

2 1944 3

3 1945 3

4 1946 1

Highest No.Of TV Shows released in the Year 2020

19 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

In [84]: # In which year highest no.of TV Schows were Released?

df_tv_y = df_tv.groupby("release_year").agg({"title":"nunique"}).reset_index()
df_tv_y.tail()

Out[84]: release_year title

41 2017 265

42 2018 380

43 2019 397

44 2020 436

45 2021 315

In [85]: #Graphical Representation using Line plot

fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
x = df_tv_y["release_year"]
y = df_tv_y["title"]
plt.plot(x,y, color ="r")
plt.title("No.Of Tv Shows released in Years")
plt.xlabel("Year")
plt.ylabel("Count Of TV Shows")

plt.subplot (2,2,2)
x = df_mov_y["release_year"]
y = df_mov_y["title"]
plt.plot(x,y, color ="g")
plt.title("No.Of Movies released in Years")
plt.xlabel("Year")
plt.ylabel("Count Of Movies")
plt.suptitle("No Of TV SHOWS & MOVIES Released ", fontsize = 24)
plt.show()

Top 5 Ratings in TV SHOWS & MOVIES

In [86]: # the ratings of the most movies and tvshows
movies = df_movies.groupby(["rating"]).agg({"title":"nunique"}).reset_index().sort_values(by=["title"],ascending=False).head(5
tvshows = df_tv.groupby(["rating"]).agg({"title":"nunique"}).reset_index().sort_values(by=["title"],ascending=False).head(5)
print(movies)
print()
print(tvshows)

rating title
8 TV-MA 2062
6 TV-14 1427
5 R 797
9 TV-PG 540
4 PG-13 490

rating title
4 TV-MA 1145
2 TV-14 733
5 TV-PG 323
7 TV-Y7 195
6 TV-Y 176

In [87]: #Graphical Representation using Bar plot

fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)

20 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

x = tvshows["rating"]
y = tvshows["title"]
height = plt.bar(x,y, color ="gray")
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("No.Of Ratings in Tv Shows")
plt.xlabel("Rating")
plt.ylabel("Count")

plt.subplot (2,2,2)
x = movies["rating"]
y = movies["title"]
height = plt.bar(x,y, color ="g")
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("No.Of Ratings in Movies")
plt.xlabel("Rating")
plt.ylabel("Count")
plt.suptitle("Top 5 Ratings in TV SHOWS & MOVIES", fontsize = 24)
plt.show()

TOP 5 Actors in Indian TV SHOWS & MOVIES

In [88]: indian_movies = df_movies[df_movies["country"]== "India"]
indian_movies_g = indian_movies.groupby("cast").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
indian_movies_g= indian_movies_g[indian_movies_g["cast"]!= "Unknown Actor"]
indian_movies_g.head()

Out[88]: cast title

0 Anupam Kher 40

1 Shah Rukh Khan 35

2 Naseeruddin Shah 32

3 Akshay Kumar 29

4 Om Puri 29

In [89]: indian_tv = df_tv[df_tv["country"]== "India"]

indian_tv_g = indian_tv.groupby("cast").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
indian_tv_g= indian_tv_g[indian_tv_g["cast"]!= "Unknown Actor"]
indian_tv_g.head(5)

Out[89]: cast title

1 Rajesh Kava 3

2 Nishka Raheja 3

3 Prakash Raj 2

4 Sabina Malik 2

5 Anjali 2

In [90]: #TOP 5 Cast in Indian TV SHOWS & MOVIES

#Graphical RepresentatioTn using bar plot
fig = plt.figure (figsize= (15,10))

21 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

plt.subplot (2,2,1)
indian_tv_t_c = indian_tv_g.head(5)
x= indian_tv_t_c["cast"]
y = indian_tv_t_c["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Actors in Indian TV SHOWS")
plt.xlabel("Top Actors")
plt.ylabel("No.of TV Shows")
plt.xticks(rotation = 15)

plt.subplot (2,2,2)
indian_mov_t_c = indian_movies_g.head(5)
x= indian_mov_t_c["cast"]
y = indian_mov_t_c["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Actors in Indian MOVIES")
plt.xlabel("Top Actors")
plt.ylabel("No.of Movies")
plt.xticks(rotation = 15)
plt.suptitle("TOP 5 Actors in Indian TV SHOWS & MOVIES", fontsize = 24)
plt.show()

TOP 5 Directors in Indian TV SHOWS & MOVIES

In [91]: # to find which indian director had more Movies
indian_movies = df_movies[df_movies["country"]== "India"]
indian_movies_g_d = indian_movies.groupby("director").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
indian_movies_g_d= indian_movies_g_d[indian_movies_g_d["director"]!= "Unknown Directors"]
indian_movies_g_d.head()

Out[91]: director title

0 Rajiv Chilaka 22

1 Suhas Kadav 16

3 David Dhawan 9

4 Umesh Mehra 8

5 Anurag Kashyap 8

In [92]: # to find which indian director had more Tv Shows

indian_tv = df_tv[df_tv["country"]== "India"]
indian_tv_g_d = indian_tv.groupby("director").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
indian_tv_g_d= indian_tv_g_d[indian_tv_g_d["director"]!= "Unknown Directors"]
indian_tv_g_d.head(5)

22 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[92]: director title

1 Gautham Vasudev Menon 2

2 Abhishek Chaubey 1

3 Sudha Kongara 1

4 Rathindran R Prasad 1

5 Sankalp Reddy 1

In [93]: #TOP 5 Directors in Indian TV SHOWS & MOVIES

#Graphical RepresentatioTn using bar plot
fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
indian_tv_t_d = indian_tv_g_d.head(5)
x= indian_tv_t_d["director"]
y = indian_tv_t_d["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Directors in Indian TV SHOWS")
plt.xlabel("Top Directors")
plt.ylabel("No.of TV Shows")
plt.xticks(rotation = 15)

plt.subplot (2,2,2)
indian_mov_t_d = indian_movies_g_d.head(5)
x= indian_mov_t_d["director"]
y = indian_mov_t_d["title"]
height = plt.bar(x, y, color = ("g","orange","tab:blue", "tab:gray", "purple"))
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("Top 5 Directors in Indian MOVIES")
plt.xlabel("Top Directors")
plt.ylabel("No.of Movies")
plt.xticks(rotation = 15)
plt.suptitle("TOP 5 Directors in Indian TV SHOWS & MOVIES", fontsize = 24)
plt.show()

TOP 10 Countries Releasing TV SHOWS & MOVIES

We can understand United States, United Kingdom, Japan, South Korea & Canada are producing More TV Shows

If we look into Movies We can see United States, India, United Kingdom, Canada, France are Producing More Movies.

In [94]: # Percentage of TV SHOWS & MOVIES Releasing by Countries

#Graphical RepresentatioTn using Pie Chart
fig = plt.figure (figsize= (18,10))
plt.subplot (2,2,1)
df_country_t = df_tv.groupby("country").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index().head
x = df_country_t["country"]
y = df_country_t["title"]
plt.pie(y, labels = x, autopct = "%1.1f%%",startangle=75, pctdistance = 0.85)
plt.title("TOP 10 Countries Releasing TV Shows")

23 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

plt.subplot (2,2,2)
df_country_m = df_movies.groupby("country").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
df_country_m= df_country_m[df_country_m["country"]!= "Unknown Country"]
df_country_m = df_country_m.head(10)
x = df_country_m["country"]
y = df_country_m["title"]
plt.pie(y,labels = x, autopct = "%1.1f%%",startangle=85, pctdistance = 0.8)
plt.title("TOP 10 Countries Releasing MOVIES")
plt.suptitle("TOP 10 Countries Releasing TV SHOWS & MOVIES", fontsize = 24)

# Top 10 Countries Releasing TV SHOWS & MOVIES

#Graphical RepresentatioTn using Bar plot
colors = ["r","g","b","c","m","y","darkorange","royalblue"]
plt.subplot (2,2,3)
df_country_t = df_tv.groupby("country").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index().head
x = df_country_t["country"]
y = df_country_t["title"]
height = plt.bar(x,y, color = colors)
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.xlabel("Top Country")
plt.ylabel("No.Of TV Shows")
plt.xticks(rotation = 90)
plt.title("TOP 10 Countries Releasing TV Shows")

plt.subplot (2,2,4)
df_country_m = df_movies.groupby("country").agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
df_country_m= df_country_m[df_country_m["country"]!= "Unknown Country"]
df_country_m = df_country_m.head(10)
x = df_country_m["country"]
y = df_country_m["title"]
height = plt.bar(x,y, color = colors)
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("TOP 10 Countries Releasing MOVIES")
plt.xlabel("Top Country")
plt.ylabel("No.of Movies")
plt.xticks(rotation = 90)
plt.suptitle("TOP 10 Countries Releasing TV SHOWS & MOVIES", fontsize = 24)
plt.show()

How many Movies/ TV Shows Released in Which Month of the Year

In [95]: df_mov1 = df_movies.groupby(["month","type"]).agg({"title":"nunique"}).reset_index()
df_mov1.head()

24 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[95]: month type title

0 1 Movie 546

1 2 Movie 382

2 3 Movie 529

3 4 Movie 550

4 5 Movie 439

In [96]: df_tv1 = df_tv.groupby(["month","type"]).agg({"title":"nunique"}).reset_index()

df_tv1.head()

Out[96]: month type title

0 1 TV Show 192

1 2 TV Show 181

2 3 TV Show 213

3 4 TV Show 215

4 5 TV Show 193

In [97]: fig = plt.figure(figsize = (15, 5))

height = plt.bar(df_mov1["month"] -0.2 , df_mov1["title"] , - 0.4, label = "Movies" )
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
height = plt.bar(df_tv1["month"] + 0.2 ,df_tv1["title"] , +0.4, label = "TV Shows")
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.legend(loc = (1,1))
plt.title("How many Movies/ TV Shows Released in Which Month")
plt.xlabel("Months")
plt.ylabel("No.of Movies/TV Shows")
plt.show()

Heat Map & Pair plots

1) we will create more numerical fields to get a Heat map and pairplot here
2) although there are no meaningful numerical features in this dataset, but we can attempt to see if we can
find something interesting in the data that we have.

In [98]: df_dur = df_movies.groupby(["duration","release_year","year","month","day"]).agg({"title":"nunique"}).sort_values(by = ["title"

df_dur.head(5)

Out[98]: duration release_year year month day title

0 53 2017 2019 7 1 6

1 54 2017 2019 7 1 6

2 64 2013 2021 7 22 5

3 122 2021 2021 8 23 4

4 97 2010 2020 1 1 3

In [99]: # Correlation Heatmap for Movies

25 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

correlation_matrix = df_dur[["release_year", "duration","year","month","day"]].corr()

plt.figure(figsize=(8, 6))
sns.heatmap(correlation_matrix, annot=True, cmap="coolwarm")
plt.title('Correlation Heatmap for Movies')
plt.show()

In [100… df_dur_tv = df_tv.groupby(["duration","release_year","year","month","day"]).agg({"title":"nunique"}).sort_values(by = ["title"

df_dur_tv.head(5)

Out[100… duration release_year year month day title

0 1 2016 2017 8 1 8

1 1 2012 2016 12 15 7

2 1 2018 2018 10 1 6

3 1 2014 2017 7 1 6

4 1 2019 2019 8 16 6

In [101… # Correlation Heatmap for TV Shows

correlation_matrix = df_dur_tv[["release_year", "duration","year","month","day"]].corr()
plt.figure(figsize=(8, 6))
sns.heatmap(correlation_matrix, annot=True, cmap="coolwarm")
plt.title('Correlation Heatmap for TV Shows')
plt.show()

26 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Pair Plot
In [102… df_dur_pair = df_final.groupby(["duration","release_year","year","month","day","type"]).agg({"title":"nunique"}).sort_values(by
df_dur_pair.head(5)

Out[102… duration release_year year month day type title

0 1 2016 2017 8 1 TV Show 8

1 1 2012 2016 12 15 TV Show 7

2 54 2017 2019 7 1 Movie 6

3 53 2017 2019 7 1 Movie 6

4 1 2019 2019 8 16 TV Show 6

In [105… # Pair plot for the Numerical columns available in the data set
sns.pairplot(df_dur_pair[["release_year", "duration","year","month","day","type"]], hue="type")
plt.show()

27 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

How has the number of movies/Tv shows released per year changed
over the last 20–30 years?
• Growth in Number of Movies and TV shows skyrocketed since 2014

• The pace of Movie releases was somewhat slow until 1996, but since 1996 ther’s been notable increase in the realese of movies and Tv shows
started out with good pace since 2005

• It is interesting to note that, TV Shows ended up in higher side on comparison with movies

In [104… titles_per_year_type = df_final.groupby(['release_year', 'type']).agg({"title": "nunique"}).reset_index()[57:]

titles_per_year_type.head()

Out[104… release_year type title

57 1991 Movie 16

58 1991 TV Show 1

59 1992 Movie 20

60 1992 TV Show 3

61 1993 Movie 24

In [106… # Plot the number of titles released per year with hue as type
plt.figure(figsize=(12, 8))
sns.lineplot(x='release_year', y='title', hue='type', data=titles_per_year_type, marker='p')
plt.xlabel('Release Year')
plt.ylabel('Number of Titles')
plt.title('Number of Movies/ TV Shows Released Per Year by Type')
plt.xticks(rotation=45)
plt.grid(True)

28 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

plt.legend(title='Type')
plt.show()

Finding Outliers in No.of Seasons in TV Shows & the Duration of

Movies
In [107… # Finding Outliers in No.of Seasons in TV Shows
df_dur = df_tv.groupby(["duration","type"]).agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index()
df_dur.head(5)

Out[107… duration type title

0 1 TV Show 1793

1 2 TV Show 425

2 3 TV Show 199

3 4 TV Show 95

4 5 TV Show 65

In [108… # Finding Outliers in movie duration in Movies

df_dur_m = df_movies.groupby(["duration","type"]).agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index
df_dur_m.head(5)

Out[108… duration type title

0 90 Movie 152

1 97 Movie 146

2 93 Movie 146

3 94 Movie 146

4 91 Movie 144

In [109… #Finding Outliers in No.of Seasons in TV Shows & the Duration of Movies
#Graphical RepresentatioTn using box plot
fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
sns.boxplot (x = "type", y = "duration", data = df_dur)
plt.title("Finding Outliers in No.of Seasons in TV Shows")
plt.xlabel("TV Shows")

29 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

plt.ylabel("No of Seasons")

plt.subplot (2,2,2)
sns.boxplot (x = "type", y = "duration", data = df_dur_m)
plt.title("Finding Outliers in the Duration of Movies")
plt.xlabel("Movies")
plt.ylabel("Duration Of Movies")
plt.suptitle("Finding Outliers in No.of Seasons in TV Shows & the Duration of Movies", fontsize = 20)
plt.show()

What is the best time to launch a TV show?

Best Month to release a TV Show is July and December

Best Month to release a Movie is July

In [110… df_mov_month = df_movies.groupby("month").agg({"title":"nunique"}).reset_index()

df_mov_month.head()

Out[110… month title

0 1 546

1 2 382

2 3 529

3 4 550

4 5 439

In [111… df_tv_month = df_tv.groupby("month").agg({"title":"nunique"}).reset_index()

df_tv_month.head()

Out[111… month title

0 1 192

1 2 181

2 3 213

3 4 215

4 5 193

In [112… #Graphical Representation using Bar plot

fig = plt.figure (figsize= (15,10))
plt.subplot (2,2,1)
x = df_tv_month["month"]
y = df_tv_month["title"]
height = plt.bar(x,y, color ="orange")
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("BEST TIME TO LAUNCH TV SHOWS")
plt.xlabel("TV Shows releasing Month")
plt.ylabel("Count")

plt.subplot (2,2,2)
x = df_mov_month["month"]

30 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

y = df_mov_month["title"]
height = plt.bar(x,y, color ="g")
for bar in height:
yval = bar.get_height()
plt.text(bar.get_x() + bar.get_width()/2, yval, str(yval), ha = "center", va = "bottom")
plt.title("BEST TIME TO LAUNCH MOVIES")
plt.xlabel("Movies releasing Month")
plt.ylabel("Count")
plt.suptitle("BEST TIME TO LAUNCH TV SHOWS & MOVIES", fontsize = 24)
plt.show()

Understanding what content is available in different countries

Top 5 Countries with Number of Titles per Top 5 Genre in Movies

In [113… # Group by country and count the number of unique titles
country_titles_m = df_movies.groupby('country').agg({"title": "nunique"}).reset_index()
top_5_countries = country_titles_m.sort_values(by='title', ascending=False).head(5)

# Filter the dataset to include only the top 5 countries

top_5_country_filter = df_movies[df_movies['country'].isin(top_5_countries['country'])]
country_genre_titles = top_5_country_filter.groupby(['country', 'listed_in']).agg({"title": "nunique"}).reset_index()
genre_counts = top_5_country_filter.groupby('listed_in')['title'].nunique().reset_index()

top_genres = genre_counts.sort_values(by='title', ascending=False).head(5)['listed_in'].tolist()

country_genre_titles_top5 = country_genre_titles[country_genre_titles['listed_in'].isin(top_genres)].sort_values(by='title', ascending

plt.figure(figsize=(15, 8))
sns.barplot(y='title', x='listed_in', hue='country', data=country_genre_titles_top5)
plt.xlabel('Genre')
plt.ylabel('Number of Titles')
plt.title('Top 5 Countries with Number of Titles per Top 5 Genre in Movies')
plt.legend(title='Country')
plt.show()

31 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Top 5 Countries with Number of Titles per Top 5 Genre in TV Shows

In [114… # Group by country and count the number of unique titles
country_titles_t = df_tv.groupby('country').agg({"title": "nunique"}).reset_index()
top_5_countries_t = country_titles_t.sort_values(by='title', ascending=False).head(5)

# Filter the dataset to include only the top 5 countries

top_5_country_filter_t = df_tv[df_tv['country'].isin(top_5_countries_t['country'])]
country_genre_titles_t = top_5_country_filter_t.groupby(['country', 'listed_in']).agg({"title": "nunique"}).reset_index()
genre_counts_t = top_5_country_filter_t.groupby('listed_in')['title'].nunique().reset_index()

top_genres_t = genre_counts_t.sort_values(by='title', ascending=False).head(5)['listed_in'].tolist()

country_genre_titles_top5_t = country_genre_titles_t[country_genre_titles_t['listed_in'].isin(top_genres_t)].sort_values(by='title'

plt.figure(figsize=(15, 8))
sns.barplot(y='title', x='listed_in', hue='country', data=country_genre_titles_top5_t)
plt.xlabel('Genre')
plt.ylabel('Number of Titles')
plt.title('Top 5 Countries with Number of Titles per Top 5 Genre in TV Shows')
plt.legend(title='Country')
plt.show()

32 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Distribution of TV SHOWS & MOVIES by Date Added in Netflix

In [115… df_t_d = df_tv.groupby("date_added").agg({"title":"nunique"}).reset_index()
df_t_d.head(5)

Out[115… date_added title

0 2008-02-04 1

1 2013-03-31 1

2 2013-08-02 1

3 2013-09-01 1

4 2013-10-08 1

In [116… df_m_d = df_movies.groupby("date_added").agg({"title":"nunique"}).reset_index()

df_m_d.head(5)

Out[116… date_added title

0 2008-01-01 1

1 2009-05-05 1

2 2009-11-18 1

3 2010-11-01 1

4 2011-05-17 1

In [117… plt.figure(figsize=(20, 7))

plt.subplot(1, 2, 1)
sns.histplot(df_t_d['date_added'], bins=15, kde=True)
plt.title('Distribution of TV shows by Date Added')
plt.xlabel('Date Added')
plt.ylabel('Frequency')

plt.subplot(1, 2, 2)
sns.histplot(df_m_d['date_added'], bins=15, kde=True)
plt.title('Distribution of Movies by Date Added')
plt.xlabel('Date Added')
plt.ylabel('Frequency')
plt.suptitle("Distribution of TV SHOWS & MOVIES by Date Added in Netflix", fontsize = 24)
plt.show()

Distribution of TV SHOWS & MOVIES by Seasons & Durations

In [118… df_t_dur = df_tv.groupby("duration").agg({"title":"nunique"}).reset_index()
df_t_dur.head(5)

33 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

Out[118… duration title

0 1 1793

1 2 425

2 3 199

3 4 95

4 5 65

In [119… df_mo_dur = df_movies.groupby("duration").agg({"title":"nunique"}).reset_index()

df_mo_dur.head(5)

Out[119… duration title

0 3 1

1 5 1

2 8 1

3 9 1

4 10 1

In [120… plt.figure(figsize=(20, 7))

plt.subplot(1, 2, 1)
sns.histplot(df_t_dur['duration'], bins=15, kde=True)
plt.title('Distribution of TV shows by Seasons')
plt.xlabel('No.Of Seasons')
plt.ylabel('Frequency')

plt.subplot(1, 2, 2)
sns.histplot(df_mo_dur['duration'], bins=15, kde=True)
plt.title('Distribution of Movies by Duration')
plt.xlabel('Movie Duration')
plt.ylabel('Frequency')
plt.suptitle("Distribution of TV SHOWS & MOVIES by Seasons & Durations", fontsize = 24)
plt.show()

Finding the Top 5 Actor & Director Pair for Movies & TV Shows
In [121… cast_director_m = df_movies.groupby(["cast","director"]).agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False)
cast_director_m= cast_director_m[(cast_director_m["director"]!= "Unknown Directors") & (cast_director_m["cast"]!= "Unknown Actors"
cast_director_m.head(5)

Out[121… cast director title

1 Rajesh Kava Rajiv Chilaka 19

2 Julie Tejwani Rajiv Chilaka 19

3 Rupa Bhimani Rajiv Chilaka 18

4 Jigna Bhardwaj Rajiv Chilaka 18

5 Vatsal Dubey Rajiv Chilaka 16

In [122… cast_director_t = df_tv.groupby(["cast","director"]).agg({"title":"nunique"}).sort_values(by = ["title"], ascending=False).reset_index

cast_director_t= cast_director_t[(cast_director_t["director"]!= "Unknown Directors") & (cast_director_t["cast"]!= "Unknown Actors"

34 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

cast_director_t.head(5)

Out[122… cast director title

796 David Attenborough Alastair Fothergill 3

971 Dave Chappelle Stan Lathan 2

1653 Sung Dong-il Shin Won-ho 2

1797 Gautham Vasudev Menon Gautham Vasudev Menon 2

1846 Anjali Gautham Vasudev Menon 2

Insights based on Non-Graphical and Visual Analysis

• Distribution of Netflix collection in Movies is 69.6% and in TV Shows it is 30.4%

• Netflix collection of Tv shows span from Year 1925 to 2021 and Movies spans from Year 1942 to 2021.

• collection of Netflix consist 8807 Titles, 36440 Actors, 4994 Directors, 42 Genre, 124 Countries and 15 Content Ratings are there.

• TOP 5 GENRE IN TV SHOWS:

1. International Movies
2. Dramas
3. Tv Comedies
4. Crime TV Shows
5. Kids’ TV

• TOP 5 GENRE IN MOVIES:

1. International Movies
2. Dramas
3. Comedies
4. Documentaries
5. Action & Adventure

• TOP 5 DIRECTOR IN TV SHOWS:

1. Ken Burns
2. Alastair Fothergill
3. Stan Lathan
4. Joe Berlinger
5. Hsu Fu-chun

• TOP 5 DIRECTOR IN MOVIES:

1. Rajiv Chilaka
2. Jan Suter
3. Raúl Campos
4. Suhas Kadav
5. Marcus Raboy

• TOP 5 ACTORS IN TV SHOWS:

1. Takahiro Sakurai
2. Yuki Kaji
3. Junichi Suwabe
4. Daisuke Ono
5. Ai Kayano

• TOP 5 ACTORS IN MOVIES:

1. Anupam Kher
2. Shah Rukh Khan
3. Naseeruddin Shah
4. Om Puri
5. Akshay Kumar

• TOP 5 Ratings IN TV SHOWS:

1. TV-MA
2. TV-14

35 of 36 05-10-2024, 07:24 pm
Netflix Case Study by Pavithran file:///C:/Users/sivap/Downloads/Python/Netflix%20Case%20Study%20by%20Pavithran.html

3. TV-PG
4. TV-Y7
5. TV-Y

• TOP 5 Ratings IN MOVIES:

1. TV-MA
2. TV-14
3. R
4. TV-PG
5. PG-13

Most of Movies Movies fall under 60–145 Min duration Range

It is observed that Netflix is More focused in Mature rated Movies / TV shows

Business Insights
It is known that R rating in india has very few titles (6)

93.52% of the content was added in the netfix in 5 years(2017–2021) and the rest 6.48% was added in 9 years (2008–2017)

Average Duration of Movies is nearly 106 Min and Tv shows is 2 seasons

Top 10 Genre covers 65.77% and the rest 32 Genre covers 34.23%

Recommendation:
• In a Tech platform like Netflix, Business metrics like DAU (Daily active users), WAU (Weekly active users) & MAU (Monthly active users) are
crucially important to look out for, and the comparison of Movie and TV series, Tv series would increase the watch time of users, in turn users
would be more likely of renewing their subscription as well as more watch time ensures higher Ad revenues.

• As seen in the business insight, a movie has a possibility of 106 minutes of watch time against a Tv show of 2 seasons

• Say 2 seasons has around 10 episodes equates to 20 episodes with each assumed duration of 40 minutes totalling to 800 minutes

• As seen in the Business Insight, Most of the country’s content potential is yet to unfold, and many regional content is witnessed to
breakthrough the popularity in recent years, so it is recommended to search for such regional content that may have the right potential to
attract more users with the help of newly introduced novelty.

• It is understood that out of selective revenue models of Netflix, Ad revenue is indeed an important one

• Countries like Japan, Greece, Italy, Germany etc are known to have higher % of old people, Netflix can study the demographics of all
countries to target the most optimal audience, in general it is undoubtedly the adult users that is targeted but, tailoring the content based on
Demographics of country might help expand the userbase.

• Since pandemic, a boom in online education is seen (also known as zoom era), Netflix can tap into this demand and enable content for kids
where entertaining education can help expand userbase under “Netflix for kids!”.

• We can interpret top Genres here as widely accepted genre and content produced under such genre would have higher probability of
popularity and acceptance by the audience.

In [107… df_final.to_csv("netflix_cleaned.csv", index = False)

36 of 36 05-10-2024, 07:24 pm