In [1]:

# 데이터셋 불러오기

import pandas as pd

df = pd.read_csv('/content/drive/MyDrive/vgames2.csv')
df.drop(['Unnamed: 0'],axis=1,inplace=True)

# 결측치 확인 및 제거

print(df.isnull().sum())
df[['Year','Genre','Publisher']].dropna(inplace=True)
df.reset_index(drop=True,inplace=True)

# 데이터 타입 확인 및 변환
print(df.dtypes)
df['NA_Sales'] = df['NA_Sales'].replace({"K":"*1e3", "M":"*1e6"}, regex=True).map(pd.eval).astype(float)
df['EU_Sales'] = df['EU_Sales'].replace({"K":"*1e3", "M":"*1e6"}, regex=True).map(pd.eval).astype(float)
df['JP_Sales'] = df['JP_Sales'].replace({"K":"*1e3", "M":"*1e6"}, regex=True).map(pd.eval).astype(float)
df['Other_Sales'] = df['Other_Sales'].replace({"K":"*1e3", "M":"*1e6"}, regex=True).map(pd.eval).astype(float)
df

Name             0
Platform         0
Year           271
Genre           50
Publisher       58
NA_Sales         0
EU_Sales         0
JP_Sales         0
Other_Sales      0
dtype: int64
Name            object
Platform        object
Year           float64
Genre           object
Publisher       object
NA_Sales        object
EU_Sales        object
JP_Sales        object
Other_Sales     object
dtype: object

/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:11: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  # This is added back by InteractiveShellApp.init_path()

Out[1]:

	Name	Platform	Year	Genre	Publisher	NA_Sales	EU_Sales	JP_Sales	Other_Sales
0	Candace Kane's Candy Factory	DS	2008.0	Action	Destineer	0.04	0.00	0.00	0.00
1	The Munchables	Wii	2009.0	Action	Namco Bandai Games	0.17	0.00	0.00	0.01
2	Otome wa Oanesama Boku ni Koi Shiteru Portable	PSP	2010.0	Adventure	Alchemist	0.00	0.00	0.02	0.00
3	Deal or No Deal: Special Edition	DS	2010.0	Misc	Zoo Games	0.04	0.00	0.00	0.00
4	Ben 10 Ultimate Alien: Cosmic Destruction	PS3	2010.0	Platform	D3Publisher	0.12	0.09	0.00	0.04
...	...	...	...	...	...	...	...	...	...
16593	Ice Age 2: The Meltdown	GC	2006.0	Platform	Vivendi Games	0.15	0.04	0.00	0.01
16594	Rainbow Islands: Revolution	PSP	2005.0	Action	Rising Star Games	0.01	0.00	0.00	0.00
16595	NBA 2K16	PS3	2015.0	Sports	Take-Two Interactive	0.44	0.19	0.03	0.13
16596	Toukiden: The Age of Demons	PSV	2013.0	Action	Tecmo Koei	0.05	0.05	0.25	0.03
16597	The King of Fighters '95	PS	1996.0	Fighting	Sony Computer Entertainment	0.00	0.00	0.16	0.01

16598 rows × 9 columns

In [2]:

!pip install plotly==4.10.0
import seaborn as sns
import matplotlib.pyplot as plt
import pandas as pd
import matplotlib as mpl
import numpy as np
import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import plotly
from plotly.graph_objs import Layout

Collecting plotly==4.10.0
  Downloading plotly-4.10.0-py2.py3-none-any.whl (13.0 MB)
     |████████████████████████████████| 13.0 MB 75 kB/s 
Requirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from plotly==4.10.0) (1.15.0)
Requirement already satisfied: retrying>=1.3.3 in /usr/local/lib/python3.7/dist-packages (from plotly==4.10.0) (1.3.3)
Installing collected packages: plotly
  Attempting uninstall: plotly
    Found existing installation: plotly 4.4.1
    Uninstalling plotly-4.4.1:
      Successfully uninstalled plotly-4.4.1
Successfully installed plotly-4.10.0

In [3]:

# 지역과 선호하는 게임 장르의 연관성

df_genre = df.drop(['Name','Platform','Year','Publisher'],axis=1)

df_genre2 = df_genre.groupby(by='Genre').sum().reset_index()
df_genre2 = df_genre2.loc[:, df_genre2.columns != 'Genre'].astype(float)

In [4]:

# 카이제곱 검정
from scipy import stats
from scipy.stats import chi2_contingency
import scipy as sp

chi2, p, dof, expected = stats.chi2_contingency(df_genre2)

print(f'게임 장르와 지역간의 카이제곱 검정 결과 P밸류가 {p}로 지역간의 장르 선호도에 차이가 있다.')

게임 장르와 지역간의 카이제곱 검정 결과 P밸류가 0.0로 지역간의 장르 선호도에 차이가 있다.

In [5]:

lab = ['Action',	'Adventure',	'Fighting',	'Misc',	'Platform',	'Puzzle',	'Racing',	'Role-Playing',	'Shooter',	'Simulation',	'Sports',	'Strategy']

fig3 = make_subplots(rows=2, cols=2,subplot_titles=("NA", "EU", "JP", "Other"),
                     specs=[[{"type": "domain"},{"type": "domain"}],[{"type": "domain"}, {"type": "domain"}]])

fig3.add_trace(go.Pie(values=df_genre2['NA_Sales'],labels = lab),
              row=1, col=1)
fig3.add_trace(go.Pie(values=df_genre2['EU_Sales'],labels = lab),
              row=1, col=2)
fig3.add_trace(go.Pie(values=df_genre2['JP_Sales'],labels = lab),
              row=2, col=1)
fig3.add_trace(go.Pie(values=df_genre2['Other_Sales'],labels = lab),
              row=2, col=2)
fig3.update_layout(height=1000, width=1200, title_text="Preference for game genre according to region")

fig3.show()

fig3.write_html("/content/sample_data/fig4.html")

In [6]:

df_genre

Out[6]:

	Genre	NA_Sales	EU_Sales	JP_Sales	Other_Sales
0	Action	0.04	0.00	0.00	0.00
1	Action	0.17	0.00	0.00	0.01
2	Adventure	0.00	0.00	0.02	0.00
3	Misc	0.04	0.00	0.00	0.00
4	Platform	0.12	0.09	0.00	0.04
...	...	...	...	...	...
16593	Platform	0.15	0.04	0.00	0.01
16594	Action	0.01	0.00	0.00	0.00
16595	Sports	0.44	0.19	0.03	0.13
16596	Action	0.05	0.05	0.25	0.03
16597	Fighting	0.00	0.00	0.16	0.01

16598 rows × 5 columns

In [7]:

# 게임시장 규모

df_market = pd.DataFrame(data=[df_genre['NA_Sales'].sum(),df_genre['EU_Sales'].sum(),df_genre['JP_Sales'].sum(),df_genre['Other_Sales'].sum()])
df_market = df_market.T

sns.set(rc = {'figure.figsize':(15,10)})

df_market.columns = ['NA','EU','JP','Other']
df_market.index = ['Amount']
fig2 = px.bar(df_market,width=500, height=500,title='MARKET SIZE')
fig2.show()
fig2.write_html("/content/sample_data/fig2.html")

In [9]:

# 연도별 게임 트렌드

df['Year'] = df['Year'].dropna().astype(int)

df = df[df.Year > 100]
df.tail()

df_year = pd.crosstab(df.Genre,df.Year)
df_year = df_year.add_prefix('').reset_index().rename_axis(None, axis=1) 

df_year['1980s'] = df_year['1980']+df_year['1981']+df_year['1982']+df_year['1983']+df_year['1984']+df_year['1985']+df_year['1986']+df_year['1987']+df_year['1988']+df_year['1989']
df_year['1990s'] = df_year['1990']+df_year['1991']+df_year['1992']+df_year['1993']+df_year['1994']+df_year['1995']+df_year['1996']+df_year['1997']+df_year['1998']+df_year['1999']
df_year['2000s'] = df_year['2000']+df_year['2001']+df_year['2002']+df_year['2003']+df_year['2004']+df_year['2005']+df_year['2006']+df_year['2007']+df_year['2008']+df_year['2009']
df_year['2010s'] = df_year['2010']+df_year['2011']+df_year['2012']+df_year['2013']+df_year['2014']+df_year['2015']+df_year['2016']+df_year['2017']
df_year['2020s'] = df_year['2020']
df_year = df_year[['Genre','1980s','1990s','2000s','2010s','2020s']]

df_year = df_year.T
df_year.columns = ['Action',	'Adventure',	'Fighting', 'Misc',	'Platform',	'Puzzle',	'Racing',	'Role-Playing',	'Shooter',	'Simulation',	'Sports',	'Strategy']
df_year = df_year.drop(df_year.index[0])

df_year.plot.bar(stacked=True)

maxs=df_year.max(axis=1)
print("연도별 최대 출고 게임량")
print(maxs)
max_game = df_year.astype(int).idxmax(axis=1)
print('연도별 최대 출고 게임\n')
print(max_game)

/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:3: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

연도별 최대 출고 게임량
1980s      66.0
1990s     303.0
2000s    1573.0
2010s    1431.0
2020s       1.0
dtype: float64
연도별 최대 출고 게임

1980s        Action
1990s        Sports
2000s        Action
2010s        Action
2020s    Simulation
dtype: object

In [10]:

df_year

Out[10]:

	Action	Adventure	Fighting	Misc	Platform	Puzzle	Racing	Role-Playing	Shooter	Simulation	Sports	Strategy
1980s	66	2	4	8	32	19	8	9	30	3	22	0
1990s	158	96	191	115	124	71	179	171	136	86	303	121
2000s	1573	630	440	1015	559	356	791	728	712	547	1397	377
2010s	1431	538	194	557	151	113	232	550	390	207	567	167
2020s	0	0	0	0	0	0	0	0	0	1	0	0

In [11]:

# 출고량이 높은 게임 분석 및 시각화

df['Total_Sales'] = df['NA_Sales'] + df['EU_Sales'] + df['JP_Sales'] + df['Other_Sales']
na_max = df.loc[df.NA_Sales.idxmax(), 'Name']
eu_max = df.loc[df.EU_Sales.idxmax(), 'Name']
jp_max = df.loc[df.JP_Sales.idxmax(), 'Name']
other_max = df.loc[df.Other_Sales.idxmax(), 'Name']
total_max = df.loc[df.Total_Sales.idxmax(), 'Name']
print('북미 최대 출고 게임 :',na_max,'\n유럽 최대 출고 게임 :',eu_max,'\n일본 최대 출고 게임 :',jp_max,'\n기타지역 최대 출고 게임 :',other_max,'\n합계 최대 출고 게임 :',total_max)

북미 최대 출고 게임 : Grand Theft Auto V 
유럽 최대 출고 게임 : FIFA 14 
일본 최대 출고 게임 : Pokemon FireRed/Pokemon LeafGreen 
기타지역 최대 출고 게임 : Fallout 3 
합계 최대 출고 게임 : Grand Theft Auto V

In [12]:

# 출고량 상위 10위 시각화
df.reset_index(drop=True,inplace=True)

labels = df.iloc[[10673,12321,567,1935,1928,1777,8353,9622,7697,6314],[0]]
label = np.array(labels['Name'].tolist())
print(labels)
fig4 = px.pie(df, values=df['Total_Sales'].nlargest(10),names=label,title='TOP 10 GAMES',width=1000, height=1000)

fig4.show()
fig4.write_html("/content/sample_data/fig3.html")

                                          Name
10673                       Grand Theft Auto V
12321                              Halo: Reach
567                              Mario Kart 64
1935                       Pokemon X/Pokemon Y
1928      The Legend of Zelda: Ocarina of Time
1777         Pokemon FireRed/Pokemon LeafGreen
8353   Pokémon Yellow: Special Pikachu Edition
9622                       Super Mario Bros. 3
7697                                   FIFA 14
6314                            Super Mario 64

In [13]:

# 휴대형 플랫폼과 거치형 플랫폼 출고량 분석

platform = df.drop_duplicates(['Platform'])
platform.reset_index(drop=True, inplace=True)
print('전체 플랫폼\n',platform['Platform'])
portable = ['DS','PSP','GBA','GB','PSV','WS','GG','3DS']
holder = ['Wii','PS3','PC','PS4','PS2','XB','X360','GC','2600','NES','SAT','N64','XOne','SNES','WiiU','DC','SCD','NG','TG16','3DO','PCFX']
print('휴대형 플랫폼\n',portable,'\n\n거치형 플랫폼\n',holder)

df_platform = df.groupby(by='Platform').sum().reset_index().drop(columns = ['Year','NA_Sales',	'EU_Sales',	'JP_Sales',	'Other_Sales'])
df_platform

df_platform_potable = pd.DataFrame(data=df_platform.iloc[[2,4,5,6,9,19,20,25]])
df_platform_potable.reset_index(drop=True,inplace=True)

df_platform_holder = df_platform.drop([2,4,5,6,9,19,20,25])
df_platform_holder.reset_index(drop=True,inplace=True)

platform_vs = [['Portable',df_platform_potable['Total_Sales'].sum()],['Holder',df_platform_holder['Total_Sales'].sum()]]

df_platform_vs = pd.DataFrame(platform_vs,columns = ['kind','Total_Sales'])
print(df_platform_vs)
fig5 = px.pie(df_platform_vs,values=df_platform_vs['Total_Sales'],names=df_platform_vs['kind'],width=600)
fig5.show()
fig5.write_html("/content/sample_data/fig5.html")

전체 플랫폼
 0       DS
1      Wii
2      PSP
3      PS3
4       PC
5       PS
6      GBA
7      PS4
8      PS2
9       XB
10    X360
11      GC
12     3DS
13    2600
14     SAT
15      GB
16     NES
17      DC
18     N64
19    XOne
20    SNES
21    WiiU
22     PSV
23     GEN
24     SCD
25      WS
26      NG
27    TG16
28     3DO
29      GG
30    PCFX
Name: Platform, dtype: object
휴대형 플랫폼
 ['DS', 'PSP', 'GBA', 'GB', 'PSV', 'WS', 'GG', '3DS'] 

거치형 플랫폼
 ['Wii', 'PS3', 'PC', 'PS4', 'PS2', 'XB', 'X360', 'GC', '2600', 'NES', 'SAT', 'N64', 'XOne', 'SNES', 'WiiU', 'DC', 'SCD', 'NG', 'TG16', '3DO', 'PCFX']
       kind   Total_Sales
0  Portable  2.936193e+07
1    Holder  1.110367e+08

In [14]:

df_platform_potable

Out[14]:

	Platform	Total_Sales
0	3DS	6620238.96
1	DS	8080792.69
2	GB	3190251.04
3	GBA	8990303.67
4	GG	0.04
5	PSP	1690286.29
6	PSV	790060.14
7	WS	1.42

In [15]:

# 휴대형 플랫폼 내 출고량 분석

portable1= [['NINTENDO',df_platform_potable.iloc[0][1]+df_platform_potable.iloc[1][1]+df_platform_potable.iloc[2][1]+df_platform_potable.iloc[3][1]],
           ['SONY',df_platform_potable.iloc[5][1]+df_platform_potable.iloc[6][1]],
           ['OTHER',df_platform_potable.iloc[4][1]+df_platform_potable.iloc[7][1]]]
df_portable = pd.DataFrame(portable1,columns=['Company','Sales'])

fig6 = make_subplots(rows=1, cols=2,subplot_titles=("Analysis by Platform","Analysis by Company"),
                     specs=[[{"type": "domain"},{"type": "domain"}]])

fig6.add_trace(go.Pie(values=df_platform_potable['Total_Sales'],labels = df_platform_potable['Platform']),
              row=1, col=1)
fig6.add_trace(go.Pie(values=df_portable['Sales'],labels = df_portable['Company']),
              row=1, col=2)

fig6.show()
fig6.write_html("/content/sample_data/fig6.html")

In [16]:

df_platform_holder

Out[16]:

	Platform	Total_Sales
0	2600	110086.37
1	3DO	0.10
2	DC	15.95
3	GC	1160195.13
4	GEN	260028.09
5	N64	14490202.43
6	NES	730249.27
7	NG	1.44
8	PC	2240251.31
9	PCFX	0.03
10	PS	14860711.02
11	PS2	12521210.92
12	PS3	12690935.21
13	PS4	3520268.01
14	SAT	1410030.93
15	SCD	1.86
16	SNES	240198.60
17	TG16	0.16
18	Wii	8760899.59
19	WiiU	1850079.94
20	X360	30550927.88
21	XB	2220248.28
22	XOne	3420137.68

In [17]:

# 거치형 플랫폼 내 출고량 분석

holder1 = [['SONY',df_platform_holder.iloc[10][1]+df_platform_holder.iloc[11][1]+df_platform_holder.iloc[12][1]+df_platform_holder.iloc[13][1]],
          ['NINTENDO',df_platform_holder.iloc[3][1]+df_platform_holder.iloc[5][1]+df_platform_holder.iloc[6][1]+df_platform_holder.iloc[16][1]+df_platform_holder.iloc[18][1]+df_platform_holder.iloc[19][1]],
          ['SEGA',df_platform_holder.iloc[2][1]+df_platform_holder.iloc[4][1]+df_platform_holder.iloc[14][1]+df_platform_holder.iloc[15][1]],
          ['MS',df_platform_holder.iloc[20][1]+df_platform_holder.iloc[21][1]+df_platform_holder.iloc[22][1]],
          ['Other',df_platform_holder.iloc[0][1]+df_platform_holder.iloc[1][1]+df_platform_holder.iloc[7][1]+df_platform_holder.iloc[8][1]+df_platform_holder.iloc[9][1]+df_platform_holder.iloc[17][1]]]

df_holder = pd.DataFrame(holder1, columns=['Company','Sales'])
df_holder

fig7 = px.pie(df_platform_holder,names=df_platform_holder['Platform'],values=df_platform_holder['Total_Sales'], height=600)
fig7.show()
fig7.write_html("/content/sample_data/fig7.html")

fig8 = px.pie(df_holder, names=df_holder['Company'],values=df_holder['Sales'], height=600)
fig8.show()
fig8.write_html("/content/sample_data/fig8.html")

In [ ]: