Complete movies visualizations
19 min readJun 12, 2021
Data Visualization is one of the fundamental skills in the Data Scientist toolkit. Data visualization is the graphical representation of information and data. By using visual elements like charts, graphs, and maps, data visualization tools provide an accessible way to see and understand trends, outliers, and patterns in data.
Table of Contents :-
Introduction :-
Dataset
- I choose the TMDb movie data set for data analysis. This data set contains information about 10,000 movies collected from The Movie Database (TMDb), including user ratings and revenue.I would like to find other intresting patterns in the dataset.
- Total Rows = 10866
- Total Columns = 21(features)
Questions :-
In this notebook we are going to analyze the movies dataset and we will find out the answers for following questions.
- Which year has the highest release of movies?
- Which Movie Has The Highest Or Lowest Profit? Top 10 movies which earn highest profit?
- Movie with Highest And Lowest Budget?
- Which movie made the highest revenue and lowest as well?
- Movie with shorest and longest runtime?
- Which movie get the highest or lowest votes (Ratings).
- Which Year Has The Highest Profit Rate?
- Which length movies most liked by the audiences according to their popularity?
- Average Runtime Of Movies From Year To Year?
- How Does The Revenue And Popularity differs Budget And Runtime? And How Does Popularity Depends On Profit?
- Which Month Released Highest Number Of Movies In All Of The Years? And Which Month Made The Highest Average Revenue?
- Which Genre Has The Highest Release Of Movies?
- Which genres are most popular from year to year?
- Most Frequent star cast?
- Top 20 Production Companies With Higher Number Of Release?
- Life Time Profit Earn By Each Production Company?
- Top 20 Director Who Directs Maximum Movies?
- What kinds of properties are associated with movies that have high revenues?
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as pltData Processing
After Observing the dataset and the questions related to this dataset for the analysis we will be keeping only relevent data and deleting the unused data.
General Properties
# Read the CSV File Using Pandas read_csv function
reader = pd.read_csv('./tmdb_movies_data.csv')
# print the concise summery of the dataset
reader.info()<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10866 entries, 0 to 10865
Data columns (total 21 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 id 10866 non-null int64
1 imdb_id 10856 non-null object
2 popularity 10866 non-null float64
3 budget 10866 non-null int64
4 revenue 10866 non-null int64
5 original_title 10866 non-null object
6 cast 10790 non-null object
7 homepage 2936 non-null object
8 director 10822 non-null object
9 tagline 8042 non-null object
10 keywords 9373 non-null object
11 overview 10862 non-null object
12 runtime 10866 non-null int64
13 genres 10843 non-null object
14 production_companies 9836 non-null object
15 release_date 10866 non-null object
16 vote_count 10866 non-null int64
17 vote_average 10866 non-null float64
18 release_year 10866 non-null int64
19 budget_adj 10866 non-null float64
20 revenue_adj 10866 non-null float64
dtypes: float64(4), int64(6), object(11)
memory usage: 1.7+ MB#print first two rows of the dataset
reader.head(2)
Observation From The Dataset
- The columns ‘budget’, ‘revenue’, ‘budget_adj’, ‘revenue_adj’ has not given currency.But for this dataset i will assume the currency is in US dollor.
#print concise summery of the dataset
reader.describe()
#since the dataset contain null values also
#count total rows in each column which contain null values
reader.isna().sum()id 0
imdb_id 10
popularity 0
budget 0
revenue 0
original_title 0
cast 76
homepage 7930
director 44
tagline 2824
keywords 1493
overview 4
runtime 0
genres 23
production_companies 1030
release_date 0
vote_count 0
vote_average 0
release_year 0
budget_adj 0
revenue_adj 0
dtype: int64
Data Cleaning
We need to perform these 4 processes.
- We need to remove duplicate rows from the dataset
- Changing format of release date into datetime format
- Remove the unused colums that are not needes in the analysis process.
- Remove the movies which are having zero value of budget and revenue.
1. Remove Duplicate Rows
#'duplicated()' function in pandas return the duplicate row as True
#for counting the duplicate elements we sum all the rows
sum(reader.duplicated())
1#After calculating the duplicate row we can drop these row using 'drop_duplicates()' function
reader.drop_duplicates(inplace = True)
#afetr removing duplicate value from the dataset
print("Afetr Removing Duplicate Values size is : ",reader.shape)Afetr Removing Duplicate Values size is : (10865, 21)
2. Changing Format Of Release Date Into Datetime Format
#the release date is in string format.
#So we need to change this in datetime format
reader['release_date'] = pd.to_datetime(reader['release_date'])
reader['release_date'].head()0 2015-06-09
1 2015-05-13
2 2015-03-18
3 2015-12-15
4 2015-04-01
Name: release_date, dtype: datetime64[ns]
3. Remove the unused colums that are not needes in the analysis process
We can see that 21 columns in the dataset, We can drop the the colums which are not usable in the data analysis process. columns like: imdb_id,overview etc.
The columns like imdb_id, homepage,tagline, overview, budget_adj and revenue_adj are not required for my analysis and I will drop these columns.
reader.drop(['budget_adj','revenue_adj','overview','imdb_id','homepage','tagline'],axis =1,inplace = True)
print("After Removing Unused Columns size is : ",reader.shape)After Removing Unused Columns size is : (10865, 15)
4. Drop theses rows which contain incorrect or inappropriate values.
There are lots of movies where the budget or revenue have a value of ‘0’ which means that the values of those variables of those movies has not been recorded. Calculating the profits of these movies would lead to inappropriate results. I think this may be due to varying factors like the lack of information, or the movies that were never released. I have chosen to eradicate these values during the data cleaning phase.
# Checking for an zero values in the budget and revenue columns
print("Rows With Zero Values In The Budget Column:",reader[(reader['budget']==0)].shape[0])
print("Rows With Zero Values In The Revenue Column:",reader[(reader['revenue']==0)].shape[0])Rows With Zero Values In The Budget Column: 5696
Rows With Zero Values In The Revenue Column: 6016
Since now we have the columns, rows and format of the dataset in right way, its time to investigate the data for the questions asked.
Exploratory Data Analysis :-
Question 1 : Which year has the highest release of movies?
# make group for each year and count the number of movies in each year
data=reader.groupby('release_year').count()['id']
print(data.tail())
#make group of the data according to their release year and count the total number of movies in each year and plot.
reader.groupby('release_year').count()['id'].plot(xticks = np.arange(1960,2016,5))
#set the figure size and labels
sns.set(rc={'figure.figsize':(10,5)})
plt.title("Year Vs Number Of Movies",fontsize = 14)
plt.xlabel('Release year',fontsize = 13)
plt.ylabel('Number Of Movies',fontsize = 13)
#set the style sheet
sns.set_style("whitegrid")release_year
2011 540
2012 588
2013 659
2014 700
2015 629
Name: id, dtype: int64
After Seeing the plot and the output we can conclude that year 2014 year has the highest release of movies followed by year 2013 .
Question 2 : Which Movie Has The Highest Or Lowest Profit?
#calculate Profit for each of the movie
#add a new column in the dataframe name 'Profit'
reader['Profit'] = reader['revenue'] - reader['budget']#use the function 'idmin' to find the index of lowest profit movie.
#use the function 'idmax' to find the index of Highest profit movie.
#print the row related to this index.
def find_minmax(x):
min_index = reader[x].idxmin()
high_index = reader[x].idxmax()
high = pd.DataFrame(reader.loc[high_index,:])
low = pd.DataFrame(reader.loc[min_index,:])
#print the movie with high and low profit
print("Movie Which Has Highest "+ x + " : ",reader['original_title'][high_index])
print("Movie Which Has Lowest "+ x + " : ",reader['original_title'][min_index])
return pd.concat([high,low],axis = 1)
#call the find_minmax function.
find_minmax('Profit')Movie Which Has Highest Profit : Avatar
Movie Which Has Lowest Profit : The Warrior's Way
The first column shows the highest profit made by a movie and second column shows the highest in loss movie in this dataset.
As we can see that ‘Avatar’ movie Directed by James Cameron earn the highest profit in all, making over 2.5B in profit in this dataset.And the most in loss movie in this dataset is The Warrior’s Way. Going in loss by more than 400M was directed by Singmoo Lee.
#make a plot which contain top 10 movies which earn highest profit.
#sort the 'Profit' column in decending order and store it in the new dataframe,
info = pd.DataFrame(reader['Profit'].sort_values(ascending = False))
info['original_title'] = reader['original_title']
data = list(map(str,(info['original_title'])))
x = list(data[:10])
y = list(info['Profit'][:10])
#make a plot usinf pointplot for top 10 profitable movies.
ax = sns.pointplot(x=y,y=x)
#setup the figure size
sns.set(rc={'figure.figsize':(10,5)})
#setup the title and labels of the plot.
ax.set_title("Top 10 Profitable Movies",fontsize = 15)
ax.set_xlabel("Profit",fontsize = 13)
sns.set_style("darkgrid")
Question 3 : Movie with Highest And Lowest Budget?
#make sure that movie with budget zero didn't affect the result.
#so change the zero into NAN in budget column.
reader['budget'] = reader['budget'].replace(0,np.NAN)
find_minmax('budget')Movie Which Has Highest budget : The Warrior's Way
Movie Which Has Lowest budget : Fear Clinic
#make a plot which contain top 10 highest budget movies.
#sort the 'budget' column in decending order and store it in the new dataframe.
info = pd.DataFrame(reader['budget'].sort_values(ascending = False))
info['original_title'] = reader['original_title']
data = list(map(str,(info['original_title'])))
#extract the top 10 budget movies data from the list and dataframe.
x = list(data[:10])
y = list(info['budget'][:10])
#plot the figure and setup the title and labels.
ax = sns.pointplot(x=y,y=x)
sns.set(rc={'figure.figsize':(10,5)})
ax.set_title("Top 10 High Budget Movies",fontsize = 15)
ax.set_xlabel("Budget",fontsize = 13)
sns.set_style("darkgrid")
Question 4 : Movie with Largest And Lowest Earned Revenue?
find_minmax('revenue')Movie Which Has Highest revenue : Avatar
Movie Which Has Lowest revenue : Wild Card
The first column shows the highest revenue made by a movie and second column shows the lowest revenue movie in this dataset. As we can see that ‘Avatar’ movie Directed by James Cameron made the highest revenue in all, making over 2.78B revenue in this dataset. And the movie which made lowest revenue is ‘Wild Card’ directed by Simon West.
#top 10 movies which made highest revenue.
#sort the 'revenue' column in decending order and store it in the new dataframe.
info = pd.DataFrame(reader['revenue'].sort_values(ascending = False))
info['original_title'] = reader['original_title']
data = list(map(str,(info['original_title'])))
#extract the top 10 movies with high revenue data from the list and dataframe.
x = list(data[:10])
y = list(info['revenue'][:10])
#make the point plot and setup the title and labels.
ax = sns.pointplot(x=y,y=x)
sns.set(rc={'figure.figsize':(10,5)})
ax.set_title("Top 10 High Revenue Movies",fontsize = 15)
ax.set_xlabel("Revenue",fontsize = 13)
sns.set_style("darkgrid")
Question 5 : Movie with Longest And Shortest Rintime?
#it can't be possible a movie with zero runtime,but dataset contain some data like this.
#so update zero with NAN in runtime column.
reader['runtime'] = reader['runtime'].replace(0,np.NAN)
find_minmax('runtime')Movie Which Has Highest runtime : The Story of Film: An Odyssey
Movie Which Has Lowest runtime : Fresh Guacamole
So again the first column shows the runtime of the highest and second the lowest with column names as the index number.
I have never heard a runtime of a movie so long, Runtime of 900 min, that’s approx 15 hrs! So ‘The Story of Film: An Odyssey’ movie has the highest runtime.
The name of the movie with shortest runtime is Fresh Guacamole, runtime of just 2 min! Woah! I have never seen such a short movie in my lifetime.
#top 10 Movies With Longest runtime
#sort the 'runtime' column in decending order and store it in the new dataframe.
info = pd.DataFrame(reader['runtime'].sort_values(ascending = False))
info['original_title'] = reader['original_title']
data = list(map(str,(info['original_title'])))
#extract the top 10 longest duraton movies data from the list and dataframe.
x = list(data[:10])
y = list(info['runtime'][:10])
#make the point plot and setup the title and labels.
ax = sns.pointplot(x=y,y=x)
sns.set(rc={'figure.figsize':(10,5)})
ax.set_title("Top 10 Longest Movies",fontsize = 15)
ax.set_xlabel("Runtime",fontsize = 13)
sns.set_style("darkgrid")
Question 6 : Movie with Highest And Lowest Votes?
find_minmax('vote_average')Movie Which Has Highest vote_average : The Story of Film: An Odyssey
Movie Which Has Lowest vote_average : Transmorphers
The first column contains the movie with highest votes and second column contain the movie with lowest votes.
As we can see that movie ‘The Story of Film: An Odyssey’ has the maximum rating (92%), which was directed by Mark Cousins and movie with lowest user ratings is ‘Transmorphers’ with 15% user ratings, which was directed by Leigh Scott.
#top 10 highets rated movies.
#sort the 'vote_average' column in decending order and store it in the new dataframe.
info = pd.DataFrame(reader['vote_average'].sort_values(ascending = False))
info['original_title'] = reader['original_title']
data = list(map(str,(info['original_title'])))
##extract the top 10 highly rated movies data from the list and dataframe.
x = list(data[:10])
y = list(info['vote_average'][:10])
#make the point plot and setup the title and labels.
ax = sns.pointplot(x=y,y=x)
sns.set(rc={'figure.figsize':(10,5)})
ax.set_title("Top 10 Highest Rated Movies",fontsize = 15)
ax.set_xlabel("Vote Average",fontsize = 13)
#setup the stylesheet
sns.set_style("darkgrid")
Question 7 : Which Year Has The Highest Profit Rate?
#year-wise analysis of profit earn by movies in each year.
#take the average of profit made by movies in each year and plot.
#make the group of the data according to their release_year and find the mean profit and plot.
reader.groupby('release_year')['Profit'].mean().plot()
#setup the title and labels of the figure.
plt.title("Year Vs Average Profit",fontsize = 14)
plt.xlabel('Release year',fontsize = 13)
plt.ylabel('Average Profit',fontsize = 13)
#setup the figure size.
sns.set(rc={'figure.figsize':(1,5)})
sns.set_style("whitegrid")
According to the plot year 2015 is most profitable years And the profit was very low between the years 1960 and 1970.
Question 8 : Which length movies most liked by the audiences according to their popularity?
#use groupby function and group the data according to their runtime.
#make a plot using their popularity and find which length movies are most popular.
#make the group of the data according to their runtime and find the mean popularity related to this and plot.
reader.groupby('runtime')['popularity'].mean().plot(figsize = (13,5),xticks=np.arange(0,1000,100))
#setup the title of the figure
plt.title("Runtime Vs Popularity",fontsize = 14)
#setup the x-label and y-label of the plot.
plt.xlabel('Runtime',fontsize = 13)
plt.ylabel('Average Popularity',fontsize = 13)
#setup the figure size.
sns.set(rc={'figure.figsize':(10,5)})
sns.set_style("whitegrid")
According to the plot we can say that movies in the range of 100–200 runtime are more popular than other runtime movies. Because it is boring to see the long duration movies.
Research Question 9: Average Runtime Of Movies From Year To Year?
#how the runtime of the movies differ year to year.
#make the group of the data according to their release_year and find the mean related to this and plot.
reader.groupby('release_year')['runtime'].mean().plot(xticks = np.arange(1960,2016,5))
#setup the figure size.
sns.set(rc={'figure.figsize':(10,5)})
#setup the title of the figure
plt.title("Runtime Vs Year",fontsize = 14)
#setup the x-label and y-label of the plot.
plt.xlabel('Year',fontsize = 13)
plt.ylabel('Runtime',fontsize = 13)
sns.set_style("whitegrid")
According to the plot movie duration is decreasing year to year and it’s TRUE. Because at this time nobody want to watch the long duration movies because it is quite boring. That’s why the average runtime duration of the movies are around 100 Minutes.
Question 10: How Does The Revenue depends upon Budget And popularity depends upon Release year And How Does Popularity Depends On Profit?
#how does revenue change according to their budget.
#make a scatter plot using 'regplot' between ''revenue' and 'budget'.
ax = sns.regplot(x=reader['revenue'], y=reader['budget'],color='c')
#set the title and labels of the figure
ax.set_title("Revenue Vs Budget",fontsize=13)
ax.set_xlabel("Revenue",fontsize=12)
ax.set_ylabel("Budget",fontsize=12)
#set the figure size
sns.set(rc={'figure.figsize':(6,4)})
sns.set_style("whitegrid")
#change zero into NAN in budget and revenue column for the exact calculation.
reader['budget'] = reader['budget'].replace(0,np.NAN)
reader['revenue'] = reader['revenue'].replace(0,np.NAN)
reader['Profit'] = reader['Profit'].replace(0,np.NAN)
#find the correlation using 'corr()' function.
#it returns a dataframe which contain the correlation between all the numeric columns.
data_corr = reader.corr()
print("Correlation Between Revenue And Budget : ",data_corr.loc['revenue','budget'])Correlation Between Revenue And Budget : 0.6885561524636727
#how does average rating change according to the release year.
#make a scatter plot using 'regplot' between 'year' and 'vote average'.
ax = sns.regplot(x=reader['release_year'], y=reader['vote_average'],color='c')
#setup the title and the labels of the scatter plot.
ax.set_title("Release Year Vs Vote Average",fontsize=13)
ax.set_xlabel("Release Year",fontsize=12)
ax.set_ylabel("Vote Average",fontsize=12)
#set the figure size and print the correlation between them.
sns.set(rc={'figure.figsize':(6,4)})
sns.set_style("whitegrid")
print("Correlation Between Release Year And Vote Average : ",data_corr.loc['release_year','vote_average'])Correlation Between Release Year And Vote Average : -0.11757558609467468
#how does profit depends upon the popularity
#make a scatter plot using 'regplot' between 'year' and 'vote average'.
ax = sns.regplot(x=reader['popularity'],y=reader['Profit'],color='c')
#setup the title and the labels of the scatter plot.
ax.set_title("Popularity Vs Profit",fontsize=13)
ax.set_xlabel("Popularity",fontsize=12)
ax.set_ylabel("Profit",fontsize=12)
#setup the figure size.
sns.set(rc={'figure.figsize':(6,4)})
sns.set_style("whitegrid")
#find the correlation between them.
print("Correlation Between Popularity And Profit : ",data_corr.loc['popularity','Profit'])Correlation Between Popularity And Profit : 0.6161404948930916
These Are Estimated Values They Can Be Differ.
- Budget vs Revenue : Budget and revenue both have positive correlation(0.68) between them. Means there is a good possibility that movies with higher investments result in better revenues.
- Release Year Vs Vote Average : Release year and vote Average have negative correlation(-0.11). Means that movie ratings(vote average) does not depends on the release year.
- Popularity Vs Profit : Popularity and profit have positive correlation(0.61). It means that movie with high popularity tends to earn high profit.
Question 11: Which Month Released Highest Number Of Movies In All Of The Years? And Which Month Made The Highest Average Revenue?
#extract the month number from the release date.
month_release = reader['release_date'].dt.month
#count the movies in each month using value_counts().
number_of_release = month_release.value_counts().sort_index()
months=['Jan','Feb','Mar','Apr','May','Jun','Jul','Aug','Sep','Oct','Nov','Dec']
number_of_release = pd.DataFrame(number_of_release)
number_of_release['month'] = months
#change the column name of the new dataframe 'number_of_release'
number_of_release.rename(columns = {'release_date':'number_of_release'},inplace=True)
#plot the bar graph using plot.
number_of_release.plot(x='month',kind='bar',fontsize = 11,figsize=(8,6))
#set the labels and titles of the plot.
plt.title('Months vs Number Of Movie Releases',fontsize = 15)
plt.xlabel('Month',fontsize = 13)
plt.ylabel('Number of movie releases',fontsize = 13)
sns.set_style("darkgrid")
According to the plot we can conclude that there are higher number of release in september and october month.
#which month made the highest average revenue?.
#make a dataframe which store the release month of each movie.
month_release = pd.DataFrame(month_release)
#change the column name of the new dataframe 'month_release'.
month_release.rename(columns = {'release_date':'release_month'},inplace=True)
#add a new column 'revenue' in the dataframe 'month_release'.
month_release['revenue'] = reader['revenue']
#make the group of the data according to their month and calculate the mean revenue of each month.
mean_revenue = month_release.groupby('release_month').mean()
mean_revenue['month'] = months
#make the bar plot using pandas plot function.
mean_revenue.plot(x='month',kind='bar',figsize = (8,6),fontsize=11)
#setup the title and lables of the plot.
plt.title('Average revenue by month (1960 - 2015)',fontsize = 15)
plt.xlabel('Month',fontsize = 13)
plt.ylabel('Average Revenue',fontsize = 13)
sns.set_style("darkgrid")
According to the plot we can say that movie which release in may or june month made the high revenue in comparison of other month release. Or it can be happened because of outliers.
Question 12: Which Genre Has The Highest Release Of Movies?
#make a function that will split the string and return the count of each genre.
def count_genre(x):
#concatenate all the rows of the genre.
data_plot = reader[x].str.cat(sep = '|')
data = pd.Series(data_plot.split('|'))
#count each of the genre and return.
info = data.value_counts(ascending=False)
return info
#call the function for counting the movies of each genre.
total_genre_movies = count_genre('genres')
#plot a 'barh' plot using plot function for 'genre vs number of movies'.
total_genre_movies.plot(kind= 'barh',figsize = (13,6),fontsize=12,colormap='tab20c')
#setup the title and the labels of the plot.
plt.title("Genre With Highest Release",fontsize=15)
plt.xlabel('Number Of Movies',fontsize=13)
plt.ylabel("Genres",fontsize= 13)
sns.set_style("whitegrid")
According to the plot Drama genre has the highest release of movies followed by Comedy and Thriller.
Question 13: Which genres are most popular from year to year?
#map all the rows of genres in a list.
genre_details = list(map(str,(reader['genres'])))
genre = ['Adventure', 'Science Fiction', 'Fantasy', 'Crime', 'Western', 'Family','nan', 'Animation','War','Mystery','Romance','TV Movie','Action', 'Thriller','Comedy','Drama' , 'History', 'Music', 'Horror', 'Documentary', 'Foreign']
#make the numpy array of year and popularity which contain all the rows of release_year and popularity column.
year = np.array(reader['release_year'])
popularity = np.array(reader['popularity'])
#make a null dataframe which indexs are genres and columns are years.
popularity_df = pd.DataFrame(index = genre, columns = range(1960, 2016))
#change all the values of the dataframe from NAN to zero.
popularity_df = popularity_df.fillna(value = 0.0)
z = 0
for i in genre_details:
split_genre = list(map(str,i.split('|')))
popularity_df.loc[split_genre, year[z]] = popularity_df.loc[split_genre, year[z]] + popularity[z]
z+=1#calculate the standard deviation for the accurate results.
def calculate_std(x):
return (x-x.mean())/x.std(ddof=0)
popular_genre = calculate_std(popularity_df)#How the popularity of the genre differ year by year.
sns.set_style("whitegrid")
fig, ax = plt.subplots(3,3,figsize = (16,10))
#set the title of the subplot.
fig.suptitle('Genre Popularity Over Year To Year',fontsize = 16)
#plot the 'Drama' genre plot see the popularity difference over year to year.
popular_genre.loc['Drama'].plot(label = "Drama",color = '#f67280',ax = ax[0][0],legend=True)
#plot the 'Action' genre plot see the popularity difference over year to year.
popular_genre.loc['Action'].plot(label = "Horror",color='#33FFB5',ax = ax[0][1],legend=True)
#plot the 'Comedy' genre plot see the popularity difference over year to year.
popular_genre.loc['Comedy'].plot(label = "Comedy",color='#fe5f55',ax = ax[0][2],legend=True)
#accordig to the previous plot the 'Thriller','Science Fiction', 'Romance', 'Music','Adventure' and 'Crime'.
popular_genre.loc['Thriller'].plot(label = "Horror",color='#00818a',ax = ax[1][0],legend=True)
popular_genre.loc['Science Fiction'].plot(label = "Science Fiction",color='#6f6600',ax = ax[1][1],legend=True)
popular_genre.loc['Romance'].plot(label = "Romance",color='#1a2c5b',ax = ax[1][2],legend=True)
popular_genre.loc['Music'].plot(label = "Music",color='#db3b61',ax = ax[2][0],legend=True)
popular_genre.loc['Adventure'].plot(label = "Adventure",color='#08c299',ax = ax[2][1],legend=True)
popular_genre.loc['Crime'].plot(label = "Crime",color='c',ax = ax[2][2],legend=True)<AxesSubplot:>
Question 14: Most Frequent Actor?
#call the 'count_genre' function which we made earlier for spiliing and counting genres.
#it wiil count number of movies done by each actor.
count_actor_flims = count_genre('cast')
#plot the barh plot.
count_actor_flims.iloc[:20].plot.bar(figsize=(13,6),colormap= 'tab20c',fontsize=12)
#setup the title and the labels of the plot.
plt.title("Most Frequent Actor",fontsize=15)
plt.xticks(rotation = 70)
plt.xlabel('Actor',fontsize=13)
plt.ylabel("Number Of Movies",fontsize= 13)
sns.set_style("whitegrid")
Question 15: Top 20 Production Companies With Higher Number Of Release?
#call the 'count_genre' function which we made earlier for spiliing and counting genres.
#it wiil count number of movies released by each production company.
production_companies = count_genre('production_companies')
#plot he barh plot.
production_companies.iloc[:20].plot(kind='barh',figsize=(16,8),fontsize=13)
plt.title("Production Companies Vs Number Of Movies",fontsize=15)
plt.xlabel('Number Of Movies',fontsize=14)
sns.set_style("whitegrid")
Question 16: Life Time Profit Earn By Each Production Company
##map all the rows of production_companies in a list.
data1 = list(map(str,(reader['production_companies'])))
production_companies = pd.DataFrame(production_companies)
#make the list of profit and production_companies which contain all the rows of profit and production_companies column.
data3 = list(production_companies.index.values)
#append the nan in the data3 because production_companies row has null values also.
data3.append('nan')
reader['Profit'] = reader['Profit'].replace(np.NAN,0)
profit = np.array(reader['Profit'])
#make a null dataframe which indexs are genres and has only one column Total Profit.
companies_df = pd.DataFrame(index = data3 ,columns =np.arange(0,1,1) )
#change all the values of the dataframe from NAN to zero.
companies_df = companies_df.fillna(value = 0.0)
z = 0
for i in data1:
split_companies = list(map(str,i.split('|')))
companies_df.loc[split_companies] = companies_df.loc[split_companies] + profit[z]
z+=1#Rename the column name
companies_df = companies_df.rename(columns={0:'Total Profit'})
#plot the barh plot
companies_df[:20].plot.barh(figsize=(13,6),fontsize=12,legend=True)
plt.title("Production Companies Vs Life Time Profit",fontsize=15)
plt.xlabel('Total Profit',fontsize=13)
sns.set_style("whitegrid")
Question 17 : Top 20 Director Who Directs Maximum Movies?
#call the 'count_genre' function which we made earlier for spiliing and counting genres.
#it wiil count number of movies directed by each diretor.
count_director_movies = count_genre('director')
#plot a bar graph
count_director_movies.iloc[:20].plot(kind='bar',figsize=(13,6),fontsize=12)
#setup the title and the labels
plt.title("Director Vs Number Of Movies",fontsize=15)
plt.xticks(rotation=70)
plt.ylabel("Number Of Movies",fontsize= 13)
sns.set_style("whitegrid")
Question 18: What Kind Of Properties Are Associated With Movies With High Revenue?
info = pd.DataFrame(reader['revenue'].sort_values(ascending=False))
data_set = ['id','popularity','budget','original_title','cast','director','runtime','genres','vote_average','release_year']
for i in data_set:
info[i] = reader[i]
info.head(3)
fig, axes = plt.subplots(2,2,figsize = (16,6))
fig.suptitle("Revenue Vs (Budget,Popularity,Vote Average,Runtime)",fontsize=14)
sns.regplot(x=reader['revenue'], y=reader['budget'],color='c',ax=axes[0][0])
sns.regplot(x=reader['revenue'], y=reader['popularity'],color='c',ax=axes[0][1])
sns.regplot(x=reader['revenue'], y=reader['vote_average'],color='c',ax=axes[1][0])
sns.regplot(x=reader['revenue'], y=reader['runtime'],color='c',ax=axes[1][1])
sns.set_style("whitegrid")
A brief description of the above plots,
Plot 1: Budget vs Revenue
There is a good possibility that movies with higher investments result in better revenues. Correlation is positive.
Plot 2: Popularity vs Revenue
The revenue seems to be increasing with popularity. We can say that if the popularity of movie is high then the revenue of the movie may be high.Correlation is positive.
Plot 3: Vote Average vs Revenue
The correlation between revenue and vote average is 0.2069. So vote average is not highly related to the revenue.
Plot 4: Runtime vs Revenue
The correlation between revenue and runtime is 0.2378. So runtime is not highly related to the revenue.
Conclusions :-
- Drama is the most popular genre, following by action, comedy and thriller.
- Drame, Comedy, Thriller and Action are top genres.
- Maximum Number Of Movies Release In year 2014.
- ‘Avatar’, ‘Star Wars’ and ‘Titanic’ are the most profitable movies.
- Short or Long duration movies are more popular than long duration movies.
- Average runtime of the movies are decreasing year by year.
- May,june,november and december are most popular month for releasing movies, if you want to earn more profit.
- Revenue is directly connected to the budget.
- Warner Bros, Universal Pictures and Paramount Pictures production companies earn more life time profit than other production companies.
- Movies with higher budgets have shown a corresponding increase in the revenues.
Limitations
- It’s not 100 percent guaranteed solution that this formula is gonna work, But it shows us that we have high probability of making high profits if we had similar characteristics as such. If we release a movie with these characteristics, it gives people high expectations from this movie. This was just one example of an influential factor that would lead to different results, there are many that have to be taken care of.
- The dataset Does not contain complete list of movies but it contains large amount of movies. some movies are also dropped which has missing information.
Thank you for reading the article . I am a movie lover and wanted to share the information about the movies to all the people like me. I have used data visualizations so all the readers can understand it without any knowledge of data science .Hope you enjoyed the article . sayonara !