This is the 22nd day of my participation in the August Wen Challenge.More challenges in August
This note is R:www.kaggle.com/umeshnaraya… Inspired by the works of the. The goal of the notes is to make it as simple as possible to implement the visualizations created in the R notebook above, using Python as well as some additional plots, and to add some comments and explanations to help Seborn/Python beginners with their data visualization/customization. We keep things interesting by playing with different colors.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
Matplotlib is building the font cache using fc-list. This may take a moment.
Copy the codeUse pandas to read in data sets. We see that each row of entries corresponds to a specific game, and the data contains the name of the game, the year it was released, and some categorical characteristics, such as platform, genre, and publisher. Finally, we see that the game (row) entry also includes the cumulative sales achieved by region, by that particular game.
df = pd.read_csv("/home/kesci/input/Datasets6073/vgsales.csv")
df.head()
Copy the code| Rank | Name | Platform | Year | Genre | Publisher | NA_Sales | EU_Sales | JP_Sales | Other_Sales | Global_Sales | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 1 | Wii Sports | Wii | 2006.0 | Sports | Nintendo | 41.49 | 29.02 | 3.77 | 8.46 | 82.74 |
| 1 | 2 | Super Mario Bros. | NES | 1985.0 | Platform | Nintendo | 29.08 | 3.58 | 6.81 | 0.77 | 40.24 |
| 2 | 3 | Mario Kart Wii | Wii | 2008.0 | Racing | Nintendo | 15.85 | 12.88 | 3.79 | 3.31 | 35.82 |
| 3 | 4 | Wii Sports Resort | Wii | 2009.0 | Sports | Nintendo | 15.75 | 11.01 | 3.28 | 2.96 | 33.00 |
| 4 | 5 | Pokemon Red/Pokemon Blue | GB | 1996.0 | Role-Playing | Nintendo | 11.27 | 8.89 | 10.22 | 1.00 | 31.37 |
Checking the maximum year value, we see that it is 2020, which is an impossible release date.
Year_data = df['Year'] print("Max Year Value: ", year_data.max()) Max Year Value: 2020.0Copy the codeBy looking at the name of the entry in the wrong year, we can search the web for the release date of the game and replace the current value with the correct release date.
max_entry = year_data.idxmax()
print(max_entry)
max_entry = df.iloc[max_entry]
pd.DataFrame(max_entry).T
5957
Copy the code| Rank | Name | Platform | Year | Genre | Publisher | NA_Sales | EU_Sales | JP_Sales | Other_Sales | Global_Sales | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 5957 | 5959 | Imagine: Makeup Artist | DS | 2020 | Simulation | Ubisoft | 0.27 | 0 | 0 | 0.02 | 0.29 |
Df ['Year'] = df['Year']. Replace (2020.0, 2009.0) print(" year_data.max()) Max Year Value: 2017.0Copy the codeNext we examine the number of games (rows) and the number of unique publishers, platforms, and genres to see how the games in our dataset are clearly distributed.
print("Number of games: ", len(df))
publishers = df['Publisher'].unique()
print("Number of publishers: ", len(publishers))
platforms = df['Platform'].unique()
print("Number of platforms: ", len(platforms))
genres = df['Genre'].unique()
print("Number of genres: ", len(genres))
Number of games: 16598
Number of publishers: 579
Number of platforms: 31
Number of genres: 12
Copy the codeLet’s do a simple null check. We might search the web for all the missing years and publishers, but now we just delete entries for games that don’t have all the data.
print(df.isnull().sum())
df = df.dropna()
Rank 0
Name 0
Platform 0
Year 271
Genre 0
Publisher 58
NA_Sales 0
EU_Sales 0
JP_Sales 0
Other_Sales 0
Global_Sales 0
dtype: int64
Copy the codeLet’s create a simple column chart to represent the total annual “worldwide sales” of video games. We get our data by data — all of our video game sales data, grouped by “year” and called.sum() to get the total for each year. This creates a data RAME with our year as the index or row name, and an entry for the total sales for that year.
In the dataset, the index for the year is a floating point number, such as “2006.0” instead of “2006.” We get our x terms by taking these values as integers. Once the data is ready, we simply pass the X and Y variables to our SebornBarart function. We also set our X label name, title, and we also rotate our XtickLabels and change their fontSize.
Y = df.groupby(['Year']).sum() y = y['Global_Sales'] x = y.index.astype(int) plt.figure(figsize=(12,8)) ax = sns.barplot(y = y, x = x) ax.set_xlabel(xlabel='$ Millions', fontsize=16) ax.set_xticklabels(labels = x, fontsize=12, rotation=50) ax.set_ylabel(ylabel='Year', fontsize=16) ax.set_title(label='Game Sales in $ Millions Per Year', fontsize=20) plt.show();Copy the code
Below we’ve created a simple column chart to represent the total number of video games released each year, but with a slight twist, it’s horizontal, which means that our year entries, usually our X-axis, are now on the Y-axis, while the count of “Global_Sales” entries, usually on the Y-axis, are now on the X-axis.
X = df.groupby(['Year']).count() x = x['Global_Sales'] y = x.index.astype(int) plt.figure(figsize=(12,8)) colors = sns.color_palette("muted") ax = sns.barplot(y = y, x = x, orient='h', palette=colors) ax.set_xlabel(xlabel='Number of releases', fontsize=16) ax.set_ylabel(ylabel='Year', fontsize=16) ax.set_title(label='Game Releases Per Year', fontsize=20)Copy the codeText(0.5, 1.0, 'Game Releases Per Year')
Copy the code
Below we have created a bullet-point chart where each publisher has the highest sales per year. Global sales are on the Y-axis and years are on the X-axis, and we use the parameter “Hue” of the pointcut diagram to represent the highest publisher.
We use a fulcrum table that makes it easy to calculate “publisher,” the publisher’s name by annual sales, and “sales,” which is the worldwide sales that the publisher generates each year.
Note that the Pivot table accepts an argument to be applied by a function with other options, such as mean, median, and mode. This pointcut requires a Dataframe, and you can simply add column names to X, Y, and Hue. We also customize our Xtick labels by rotating and resizing them.
table = df.pivot_table('Global_Sales', index='Publisher', columns='Year', aggfunc='sum') publishers = table.idxmax() sales = table.max() years = table.columns.astype(int) data = pd.concat([publishers, sales], axis=1) data.columns = ['Publisher', Figure (figsize=(12,8)) ax = sns.pointplot(y =' Global Sales', x = years, hue='Publisher', data=data, size=15) ax.set_xlabel(xlabel='Year', fontsize=16) ax.set_ylabel(ylabel='Global Sales Per Year', fontsize=16) ax.set_title(label='Highest Publisher Revenue in $ Millions Per Year', fontsize=20) ax.set_xticklabels(labels = years, fontsize=12, rotation=50) plt.show();Copy the code
Below, we create a game that generates global sales and makes the most money each year. We also returned the following data for reference. You can map different colors for each game, but adding a legend to a plot with so many purposes can make a plot look confusing.
The data creation for this graph is similar to the above, excluding the use of hues to represent categories in the data. Instead, we use a palette and pass it the numbers of colors we want from that particular palette.
table = df.pivot_table('Global_Sales', index='Name', columns='Year') table.columns = table.columns.astype(int) games = table.idxmax() sales = table.max() years = table.columns data = pd.concat([games, sales], axis=1) data.columns = ['Game', 'Global Sales'] colors = sns.color_palette("GnBu_d", Figure (figsize=(12,8) ax = SNS. Barplot (y = years, x =' Global Sales', data=data, Orient ='h', palette=colors) ax.set_xlabel(xlabel='Global Sales Per Year', fontsize=16) ax.set_ylabel(ylabel='Year', fontsize=16) ax.set_title(label='Highest Revenue Per Game in $ Millions Per Year', fontsize=20) plt.show(); dataCopy the code
| Game | Global Sales | |
|---|---|---|
| Year | ||
| 1980 | Asteroids | 4.310 |
| 1981 | Pitfall! | 4.500 |
| 1982 | Pac-Man | 7.810 |
| 1983 | Baseball | 3.200 |
| 1984 | Duck Hunt | 28.310 |
| 1985 | Super Mario Bros. | 40.240 |
| 1986 | The Legend of Zelda | 6.510 |
| 1987 | Zelda II: The Adventure of Link | 4.380 |
| 1988 | Super Mario Bros. 3 | 17.280 |
| 1989 | Tetris | 30.260 |
| 1990 | Super Mario World | 20.610 |
| 1991 | The Legend of Zelda: A Link to the Past | 4.610 |
| 1992 | Super Mario Land 2: 6 Golden Coins | 11.180 |
| 1993 | Super Mario All-Stars | 10.550 |
| 1994 | Donkey Kong Country | 9.300 |
| 1995 | Donkey Kong Country 2: Diddy’s Kong Quest | 5.150 |
| 1996 | Pokemon Red/Pokemon Blue | 31.370 |
| 1997 | Gran Turismo | 10.950 |
| 1998 | Pokémon Yellow: Special Pikachu Edition | 14.640 |
| 1999 | Pokemon Gold/Pokemon Silver | 23.100 |
| 2000 | Pokémon Crystal Version | 6.390 |
| 2001 | Gran Turismo 3: A-Spec | 14.980 |
| 2002 | Grand Theft Auto: Vice City | 16.150 |
| 2003 | Mario Kart: Double Dash!! | 6.950 |
| 2004 | Grand Theft Auto: San Andreas | 20.810 |
| 2005 | Nintendogs | 24.760 |
| 2006 | Wii Sports | 82.740 |
| 2007 | Wii Fit | 22.720 |
| 2008 | Mario Kart Wii | 35.820 |
| 2009 | Wii Sports Resort | 33.000 |
| 2010 | Kinect Adventures! | 21.820 |
| 2011 | Mario Kart 7 | 12.210 |
| 2012 | New Super Mario Bros. 2 | 9.820 |
| 2013 | Grand Theft Auto V | 18.890 |
| 2014 | Pokemon Omega Ruby/Pokemon Alpha Sapphire | 11.330 |
| 2015 | Call of Duty: Black Ops 3 | 5.064 |
| 2016 | Uncharted 4: A Thief’s End | 4.200 |
| 2017 | Phantasy Star Online 2 Episode 4: Deluxe Package | 0.020 |
data = df.groupby(['Publisher']).count().iloc[:,0] data = pd.DataFrame(data.sort_values(ascending=False))[0:10] publishers = data.index data.columns = ['Releases'] colors = sns.color_palette("spring", Len (data)) plt.figure(figsize=(12,8)) ax = sns.barplot(y = publishers, x =' Releases', data=data, Orient ='h', palette=colors) ax.set_xlabel(xlabel='Number of Releases', fontsize=16) ax.set_ylabel(ylabel='Publisher', fontsize=16) ax.set_title(label='Top 10 Total Publisher Games Released', fontsize=20) ax.set_yticklabels(labels = publishers, fontsize=14) plt.show();Copy the code
data = df.groupby(['Publisher']).sum()['Global_Sales'] data = pd.DataFrame(data.sort_values(ascending=False))[0:10] publishers = data.index data.columns = ['Global Sales'] colors = sns.color_palette("cool", Figure (figure size=(12,8)) ax = SNS. Barplot (y = publishers, x =' Global Sales', data=data, Orient ='h', palette=colors) ax.set_xlabel(xlabel='Revenue in $ Millions', fontsize=16) ax.set_ylabel(ylabel='Publisher', fontsize=16) ax.set_title(label='Top 10 Total Publisher Game Revenue', fontsize=20) ax.set_yticklabels(labels = publishers, fontsize=14) plt.show();Copy the code
rel = df.groupby(['Genre']).count().iloc[:,0] rel = pd.DataFrame(rel.sort_values(ascending=False)) genres = rel.index rel.columns = ['Releases'] colors = sns.color_palette("summer", Len (rel) plt.figure(figsize=(12,8)) ax = SNS. Barplot (y = genres, x =' Releases', data=rel, Orient ='h', palette=colors) ax.set_xlabel(xlabel='Number of Releases', fontsize=16) ax.set_ylabel(ylabel='Genre', fontsize=16) ax.set_title(label='Genres by Total Number of Games Released', fontsize=20) ax.set_yticklabels(labels = genres, fontsize=14) plt.show();Copy the code
rev = df.groupby(['Genre']).sum()['Global_Sales'] rev = pd.DataFrame(rev.sort_values(ascending=False)) genres = rev.index rev.columns = ['Revenue'] colors = sns.color_palette('Set3', Len (rev) plt.figure(figsize=(12,8) ax = SNS. Barplot (y = genres, x =' Revenue', data=rev, Orient ='h', palette=colors) ax.set_xlabel(xlabel='Revenue in $ Millions', fontsize=16) ax.set_ylabel(ylabel='Genre', fontsize=16) ax.set_title(label='Genres by Total Revenue Generated in $ Millions', fontsize=20) ax.set_yticklabels(labels = genres, fontsize=14) plt.show();Copy the code
I am white and white I, a program yuan like to share knowledge ❤️
If you don’t know how to program or want to learn, you can leave a message directly to me. Thank you very much for your likes, favorites, comments, one-click support.