In [35]:

    

from __future__ import print_function,division
import pandas as pd
import numpy as np


    

In [3]:

    

df = pd.read_csv("../../data/steam/data.clean.csv")
df.shape,df.columns


    

    
Out[3]:





((8520, 6),
 Index([u'appid', u'name', u'detailed_description', u'about_the_game',
        u'categories', u'genres'],
       dtype='object'))

In [5]:

    

df['num_labels'] = df['genres'].apply(lambda txt: len(txt.split(',')))
df


    

    
Out[5]:






  

    

      

      
appid
      
name
      
detailed_description
      
about_the_game
      
categories
      
genres
      
num_labels
    
  
  

    

      
0
      
30
      
Day of Defeat
      
enlist in an intense brand of axis vs. allied ...
      
enlist in an intense brand of axis vs. allied ...
      
Multi-player,Valve Anti-Cheat enabled
      
Action
      
1
    
    

      
1
      
40
      
Deathmatch Classic
      
enjoy fast-paced multiplayer gaming with death...
      
enjoy fast-paced multiplayer gaming with death...
      
Multi-player,Valve Anti-Cheat enabled
      
Action
      
1
    
    

      
2
      
50
      
Half-Life: Opposing Force
      
return to the black mesa research facility as ...
      
return to the black mesa research facility as ...
      
Single-player,Multi-player,Valve Anti-Cheat en...
      
Action
      
1
    
    

      
3
      
60
      
Ricochet
      
a futuristic action game that challenges your ...
      
a futuristic action game that challenges your ...
      
Multi-player,Valve Anti-Cheat enabled
      
Action
      
1
    
    

      
4
      
70
      
Half-Life
      
named game of the year by over 50 publications...
      
named game of the year by over 50 publications...
      
Single-player,Multi-player,Valve Anti-Cheat en...
      
Action
      
1
    
    

      
5
      
80
      
Counter-Strike: Condition Zero
      
with its extensive tour of duty campaign, a ne...
      
with its extensive tour of duty campaign, a ne...
      
Single-player,Multi-player,Valve Anti-Cheat en...
      
Action
      
1
    
    

      
6
      
130
      
Half-Life: Blue Shift
      
made by gearbox software and originally releas...
      
made by gearbox software and originally releas...
      
Single-player
      
Action
      
1
    
    

      
7
      
220
      
Half-Life 2
      
1998. half-life sends a shock through the game...
      
1998. half-life sends a shock through the game...
      
Single-player,Steam Achievements,Steam Trading...
      
Action
      
1
    
    

      
8
      
240
      
Counter-Strike: Source
      
the next installment of the world's # 1 online...
      
the next installment of the world's # 1 online...
      
Multi-player,Cross-Platform Multiplayer,Steam ...
      
Action
      
1
    
    

      
9
      
280
      
Half-Life: Source
      
winner of over 50 game of the year awards, hal...
      
winner of over 50 game of the year awards, hal...
      
Single-player
      
Action
      
1
    
    

      
10
      
300
      
Day of Defeat: Source
      
day of defeat offers intense online action gam...
      
day of defeat offers intense online action gam...
      
Multi-player,Cross-Platform Multiplayer,Steam ...
      
Action
      
1
    
    

      
11
      
320
      
Half-Life 2: Deathmatch
      
fast multiplayer action set in the half-life 2...
      
fast multiplayer action set in the half-life 2...
      
Multi-player,Valve Anti-Cheat enabled,Includes...
      
Action
      
1
    
    

      
12
      
340
      
Half-Life 2: Lost Coast
      
originally planned as a section of the highway...
      
originally planned as a section of the highway...
      
Single-player,Commentary available
      
Action
      
1
    
    

      
13
      
360
      
Half-Life Deathmatch: Source
      
half-life deathmatch: source is a recreation o...
      
half-life deathmatch: source is a recreation o...
      
Multi-player,Valve Anti-Cheat enabled
      
Action
      
1
    
    

      
14
      
380
      
Half-Life 2: Episode One
      
half-life 2 has sold over 4 million copies wor...
      
half-life 2 has sold over 4 million copies wor...
      
Single-player,Steam Achievements,Captions avai...
      
Action
      
1
    
    

      
15
      
400
      
Portal
      
portal is a new single player game from valve....
      
portal is a new single player game from valve....
      
Single-player,Steam Achievements,Captions avai...
      
Action
      
1
    
    

      
16
      
420
      
Half-Life 2: Episode Two
      
half-life 2: episode two is the second in a tr...
      
half-life 2: episode two is the second in a tr...
      
Single-player,Steam Achievements,Captions avai...
      
Action
      
1
    
    

      
17
      
440
      
Team Fortress 2
      
the tough break update is now available!about ...
      
"the most fun you can have online" - pc gamer ...
      
Multi-player,Cross-Platform Multiplayer,Steam ...
      
Action,Free to Play
      
2
    
    

      
18
      
500
      
Left 4 Dead
      
steam big pictureabout the gamefrom valve (the...
      
from valve (the creators of counter-strike, ha...
      
Single-player,Multi-player,Co-op,Steam Achieve...
      
Action
      
1
    
    

      
19
      
550
      
Left 4 Dead 2
      
set in the zombie apocalypse, left 4 dead 2 (l...
      
set in the zombie apocalypse, left 4 dead 2 (l...
      
Single-player,Multi-player,Co-op,Steam Achieve...
      
Action
      
1
    
    

      
20
      
570
      
Dota 2
      
dota is a competitive game of action and strat...
      
dota is a competitive game of action and strat...
      
Multi-player,Co-op,Steam Trading Cards,Steam W...
      
Action,Free to Play,Strategy
      
3
    
    

      
21
      
620
      
Portal 2
      
portal 2 draws from the award-winning formula ...
      
portal 2 draws from the award-winning formula ...
      
Single-player,Co-op,Steam Achievements,Full co...
      
Action,Adventure
      
2
    
    

      
22
      
630
      
Alien Swarm
      
alien swarm is a game and source sdk release f...
      
alien swarm is a game and source sdk release f...
      
Single-player,Multi-player,Co-op,Steam Achieve...
      
Action
      
1
    
    

      
23
      
730
      
Counter-Strike: Global Offensive
      
counter-strike: global offensive (cs: go) will...
      
counter-strike: global offensive (cs: go) will...
      
Multi-player,Steam Achievements,Full controlle...
      
Action
      
1
    
    

      
24
      
1002
      
Rag Doll Kung Fu
      
featuring a wide collection of single and mult...
      
featuring a wide collection of single and mult...
      
Single-player,Multi-player
      
Indie
      
1
    
    

      
25
      
1200
      
Red Orchestra: Ostfront 41-45
      
fight in the theatre of war that changed the w...
      
fight in the theatre of war that changed the w...
      
Multi-player,Steam Achievements,Valve Anti-Che...
      
Action
      
1
    
    

      
26
      
1250
      
Killing Floor
      
steam halloween sale - kf and all dlc - 75% of...
      
killing floor is a co-op survival horror fps s...
      
Single-player,Multi-player,Cross-Platform Mult...
      
Action
      
1
    
    

      
27
      
1300
      
SiN Episodes: Emergence
      
you are john blade, commander of hardcorps, an...
      
you are john blade, commander of hardcorps, an...
      
Single-player,Stats
      
Action
      
1
    
    

      
28
      
1500
      
Darwinia
      
combining fast-paced action with strategic bat...
      
combining fast-paced action with strategic bat...
      
Single-player
      
Indie,Strategy
      
2
    
    

      
29
      
1510
      
Uplink
      
you play an uplink agent who makes a living by...
      
you play an uplink agent who makes a living by...
      
Single-player
      
Indie,Strategy
      
2
    
    

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

      
8490
      
457230
      
Atlas Reactor VR Character Viewer
      
outwit to outlive in atlas reactor, the genre-...
      
outwit to outlive in atlas reactor, the genre-...
      
Single-player
      
Strategy
      
1
    
    

      
8491
      
457420
      
Surgeon Simulator VR: Meet The Medic
      
surgeon simulator vr: meet the medic be the me...
      
surgeon simulator vr: meet the medic be the me...
      
Single-player
      
Action,Free to Play,Indie,Simulation
      
4
    
    

      
8492
      
457440
      
Watch This!
      
watch this! is a first-person single player pl...
      
watch this! is a first-person single player pl...
      
Single-player,Steam Achievements,Partial Contr...
      
Action,Adventure,Indie
      
3
    
    

      
8493
      
457450
      
Defend your Crypt
      
defend your crypt is a strategy and puzzle gam...
      
defend your crypt is a strategy and puzzle gam...
      
Single-player,Steam Achievements,Steam Trading...
      
Indie,Strategy
      
2
    
    

      
8494
      
457480
      
Domino Sky
      
domino sky is a physics based game where you p...
      
domino sky is a physics based game where you p...
      
Single-player,Steam Achievements,Steam Trading...
      
Casual,Indie,Simulation,Strategy
      
4
    
    

      
8495
      
457490
      
Forgotten, Not Lost - A Kinetic Novel
      
an old farmer lives with his wife - however, h...
      
an old farmer lives with his wife - however, h...
      
Single-player,Steam Trading Cards
      
Casual,RPG,Simulation
      
3
    
    

      
8496
      
457520
      
Moustache Mountain
      
it is said, that an ancient civilization once ...
      
it is said, that an ancient civilization once ...
      
Single-player,Steam Achievements,Full controll...
      
Action,Casual,Indie
      
3
    
    

      
8497
      
457530
      
My Lady
      
you are miss bauxmont, the heiress to the baux...
      
you are miss bauxmont, the heiress to the baux...
      
Single-player,Steam Achievements,Captions avai...
      
Casual,Indie,Simulation
      
3
    
    

      
8498
      
457570
      
Camp Sunshine
      
camp sunshine is a 16-bit blood-soaked horror ...
      
camp sunshine is a 16-bit blood-soaked horror ...
      
Single-player,Steam Achievements,Partial Contr...
      
Action,Adventure,Casual,Indie,RPG
      
5
    
    

      
8499
      
457580
      
The Visitor
      
the visitor is a vr only horror experience abo...
      
the visitor is a vr only horror experience abo...
      
Single-player
      
Casual,Indie
      
2
    
    

      
8500
      
457690
      
Hotel Blind
      
hotel blind is a simulator of a blind person i...
      
hotel blind is a simulator of a blind person i...
      
Single-player,Steam Achievements,Partial Contr...
      
Casual,Indie,Simulation
      
3
    
    

      
8501
      
457710
      
Road Madness
      
in the 21st century, the world is occurred by ...
      
in the 21st century, the world is occurred by ...
      
Single-player,Steam Achievements,Steam Trading...
      
Action,Racing
      
2
    
    

      
8502
      
457790
      
Capria: Magic of the Elements
      
capria: magic of the elements is a first-perso...
      
capria: magic of the elements is a first-perso...
      
Single-player
      
Action,Casual,Indie,Early Access
      
4
    
    

      
8503
      
457820
      
Outrage
      
outrage is a short cyberpunk dungeon crawler. ...
      
outrage is a short cyberpunk dungeon crawler. ...
      
Single-player,Steam Achievements,Steam Trading...
      
Adventure,Indie,RPG
      
3
    
    

      
8504
      
457860
      
Apollo 11 VR
      
apollo 11 vr is the story of the greatest jour...
      
apollo 11 vr is the story of the greatest jour...
      
Single-player
      
Adventure,Simulation
      
2
    
    

      
8505
      
457870
      
Minigame Party VR
      
what is minigame party vr?minigame party vr is...
      
what is minigame party vr?minigame party vr is...
      
Single-player
      
Action,Casual,Indie
      
3
    
    

      
8506
      
457930
      
Starship: Nova Strike
      
we have been waiting for you commander. the dz...
      
we have been waiting for you commander. the dz...
      
Single-player
      
Action,Casual,Indie
      
3
    
    

      
8507
      
457940
      
Krog Wars
      
a new action game that can be played in 2d mod...
      
a new action game that can be played in 2d mod...
      
Single-player
      
Action,Casual,Indie
      
3
    
    

      
8508
      
457960
      
Holopoint
      
holopoint is pure archery madness. fight your ...
      
holopoint is pure archery madness. fight your ...
      
Single-player,Full controller support
      
Action,Indie,Simulation,Sports
      
4
    
    

      
8509
      
458030
      
StarFringe: Adversus
      
this is the story about the beginning of the c...
      
this is the story about the beginning of the c...
      
Single-player,Steam Achievements
      
Strategy,Early Access
      
2
    
    

      
8510
      
458290
      
Space Bit Attack
      
notice: made for vrspace bit attack was made w...
      
the galaxy needs you! transport yourself into ...
      
Single-player,Full controller support
      
Action,Indie
      
2
    
    

      
8511
      
458370
      
VR Baseball - Home Run Derby
      
there is no feeling greater than stepping up t...
      
there is no feeling greater than stepping up t...
      
Single-player,Full controller support,VR Support
      
Indie,Sports
      
2
    
    

      
8512
      
458420
      
Last Heroes 3
      
presentationlast heroes is an rpg developed by...
      
presentationlast heroes is an rpg developed by...
      
Single-player,Full controller support,Steam Tr...
      
Adventure,Casual,Indie,RPG,Strategy
      
5
    
    

      
8513
      
458700
      
Cursor Challenge
      
do you want to test your skills and your refle...
      
do you want to test your skills and your refle...
      
Single-player,Co-op
      
Casual,Indie
      
2
    
    

      
8514
      
458900
      
Julai
      
julai - is an arcade flight shooter.the action...
      
julai - is an arcade flight shooter.the action...
      
Single-player
      
Action,Casual,Indie,Simulation
      
4
    
    

      
8515
      
459100
      
Epsilon corp.
      
epsilon is a secret community. there are a lot...
      
epsilon is a secret community. there are a lot...
      
Single-player
      
Action,Adventure,Indie
      
3
    
    

      
8516
      
459260
      
Bowslinger
      
disclaimer: this game requires an htc vive to ...
      
disclaimer: this game requires an htc vive to ...
      
Single-player
      
Casual,Indie,Sports
      
3
    
    

      
8517
      
459310
      
The Hero Project: Redemption Season
      
america's #1 reality show for heroes is back f...
      
america's #1 reality show for heroes is back f...
      
Single-player,Steam Achievements,Captions avai...
      
Adventure,Indie,RPG
      
3
    
    

      
8518
      
459630
      
R.C. Bot Inc.
      
energy and money: iridium. the race has begun!...
      
energy and money: iridium. the race has begun!...
      
Single-player,Steam Achievements,Full controll...
      
Action,Casual,Indie,Simulation,Strategy
      
5
    
    

      
8519
      
460150
      
One Last Chance
      
do you remember that high school crush you nev...
      
do you remember that high school crush you nev...
      
Single-player
      
Casual,Indie,Simulation
      
3
    
  

8520 rows × 7 columns

In [14]:

    

# label cardinality
cardinality = df['num_labels'].mean()
cardinality


    

    
Out[14]:





2.4605633802816902

In [41]:

    

df['labelset'] = df['genres'].map(lambda txt: set(txt.split(',')))

label_list_list = df['labels'].tolist()
flattened_list_of_labels = [label for label_list in label_list_list for label in label_list]
unique_labels = set(flattened_list_of_labels)
L = len(unique_labels)
L,unique_labels


    

    
Out[41]:





(23,
 {'Accounting',
  'Action',
  'Adventure',
  'Animation & Modeling',
  'Audio Production',
  'Casual',
  'Design & Illustration',
  'Early Access',
  'Education',
  'Free to Play',
  'Indie',
  'Massively Multiplayer',
  'Photo Editing',
  'RPG',
  'Racing',
  'Simulation',
  'Software Training',
  'Sports',
  'Strategy',
  'Uncategorized',
  'Utilities',
  'Video Production',
  'Web Publishing'})

In [42]:

    

# label density
label_density = cardinality / L
label_density


    

    
Out[42]:





0.10698101653398653

In [55]:

    

# number of distinct labelsets
distinct_labelsets = []

for labelset in df['labelset'].tolist():
    
    x = str(sorted(tuple(labelset)))
    
    if x not in distinct_labelsets:
        distinct_labelsets.append(x)

num_distinct_labelsets = len(distinct_labelsets)
num_distinct_labelsets


    

    
Out[55]:





585

In [56]:

    

# number of labelsets appearing in a single instance

# this code is far from optimal; crude quick solution only

occurrences = dict()
for x in distinct_labelsets:
    occurrences[x] = 0
    
for index,row in df.iterrows():
    labelset_str = str(sorted(tuple(row['labelset'])))
    occurrences[labelset_str] += 1
    
one_hit_wonders = 0    
    
for k,v in occurrences.iteritems():
    if v == 1:
        one_hit_wonders += 1
        
one_hit_wonders


    

    
Out[56]:





237