Link Search Menu Expand Document
Machine Learning

Pandas GroupBy Operations

Understanding GroupBy objects 

import pandas as pd

titanic = pd.read_csv("titanic.csv")

titanic.head()
survived pclass sex age sibsp parch fare embarked deck
0 0 3 male 22.0 1 0 7.2500 S NaN
1 1 1 female 38.0 1 0 71.2833 C C
2 1 3 female 26.0 0 0 7.9250 S NaN
3 1 1 female 35.0 1 0 53.1000 S C
4 0 3 male 35.0 0 0 8.0500 S NaN 
titanic.tail()

titanic.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 891 entries, 0 to 890
Data columns (total 9 columns):
 #   Column    Non-Null Count  Dtype  
---  ------    --------------  -----  
 0   survived  891 non-null    int64  
 1   pclass    891 non-null    int64  
 2   sex       891 non-null    object 
 3   age       714 non-null    float64
 4   sibsp     891 non-null    int64  
 5   parch     891 non-null    int64  
 6   fare      891 non-null    float64
 7   embarked  889 non-null    object 
 8   deck      203 non-null    object 
dtypes: float64(2), int64(4), object(3)
memory usage: 62.8+ KB

titanic_slice = titanic.iloc[:10, [2,3]]

titanic_slice
sex age
0 male 22.0
1 female 38.0
2 female 26.0
3 female 35.0
4 male 35.0
5 male NaN
6 male 54.0
7 male 2.0
8 female 27.0
9 female 14.0 
titanic_slice.groupby("sex")

<pandas.core.groupby.generic.DataFrameGroupBy object at 0x0000027661634B20>

gbo = titanic_slice.groupby("sex")

type(gbo)

pandas.core.groupby.generic.DataFrameGroupBy

gbo.groups

{'female': [1, 2, 3, 8, 9], 'male': [0, 4, 5, 6, 7]}

l = list(gbo)

l

[('female',
        sex   age
  1  female  38.0
  2  female  26.0
  3  female  35.0
  8  female  27.0
  9  female  14.0),
 ('male',
      sex   age
  0  male  22.0
  4  male  35.0
  5  male   NaN
  6  male  54.0
  7  male   2.0)]

len(l)

2

l[0]

('female',
       sex   age
 1  female  38.0
 2  female  26.0
 3  female  35.0
 8  female  27.0
 9  female  14.0)

type(l[0])

tuple

l[0][0]

'female'

l[0][1]
sex age
1 female 38.0
2 female 26.0
3 female 35.0
8 female 27.0
9 female 14.0 
type(l[0][1])

pandas.core.frame.DataFrame

l[1]

('male',
     sex   age
 0  male  22.0
 4  male  35.0
 5  male   NaN
 6  male  54.0
 7  male   2.0)

titanic_slice.loc[titanic_slice.sex == "female"]
sex age
1 female 38.0
2 female 26.0
3 female 35.0
8 female 27.0
9 female 14.0 
titanic_slice_f = titanic_slice.loc[titanic_slice.sex == "female"]
titanic_slice_f
sex age
1 female 38.0
2 female 26.0
3 female 35.0
8 female 27.0
9 female 14.0 
titanic_slice_m = titanic_slice.loc[titanic_slice.sex == "male"]
titanic_slice_m
sex age
0 male 22.0
4 male 35.0
5 male NaN
6 male 54.0
7 male 2.0 
titanic_slice_f.equals(l[0][1])

True

for element in gbo:
    print(element[1])

      sex   age
1  female  38.0
2  female  26.0
3  female  35.0
8  female  27.0
9  female  14.0
    sex   age
0  male  22.0
4  male  35.0
5  male   NaN
6  male  54.0
7  male   2.0

Splitting with many Keys 

import pandas as pd

summer = pd.read_csv("datasets/summer.csv")

summer.head()
Year City Sport Discipline Athlete Country Gender Event Medal
0 1896 Athens Aquatics Swimming HAJOS, Alfred HUN Men 100M Freestyle Gold
1 1896 Athens Aquatics Swimming HERSCHMANN, Otto AUT Men 100M Freestyle Silver
2 1896 Athens Aquatics Swimming DRIVAS, Dimitrios GRE Men 100M Freestyle For Sailors Bronze
3 1896 Athens Aquatics Swimming MALOKINIS, Ioannis GRE Men 100M Freestyle For Sailors Gold
4 1896 Athens Aquatics Swimming CHASAPIS, Spiridon GRE Men 100M Freestyle For Sailors Silver 
summer.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 31165 entries, 0 to 31164
Data columns (total 9 columns):
 #   Column      Non-Null Count  Dtype 
---  ------      --------------  ----- 
 0   Year        31165 non-null  int64 
 1   City        31165 non-null  object
 2   Sport       31165 non-null  object
 3   Discipline  31165 non-null  object
 4   Athlete     31165 non-null  object
 5   Country     31161 non-null  object
 6   Gender      31165 non-null  object
 7   Event       31165 non-null  object
 8   Medal       31165 non-null  object
dtypes: int64(1), object(8)
memory usage: 2.1+ MB

summer.Country.nunique()

147

split1 = summer.groupby("Country")

l = list(split1)
l[:2]

[('AFG',
         Year     City      Sport Discipline           Athlete Country Gender  \
  28965  2008  Beijing  Taekwondo  Taekwondo  NIKPAI, Rohullah     AFG    Men   
  30929  2012   London  Taekwondo  Taekwondo  NIKPAI, Rohullah     AFG    Men   
  
              Event   Medal  
  28965     - 58 KG  Bronze  
  30929  58 - 68 KG  Bronze  ),
 ('AHO',
         Year   City    Sport Discipline          Athlete Country Gender  \
  19323  1988  Seoul  Sailing    Sailing  BOERSMA, Jan D.     AHO    Men   
  
                       Event   Medal  
  19323  Board (Division Ii)  Silver  )]

len(l)

147

l[100][1]
Year City Sport Discipline Athlete Country Gender Event Medal
5031 1928 Amsterdam Aquatics Swimming YLDEFONSO, Teofilo PHI Men 200M Breaststroke Bronze
5741 1932 Los Angeles Aquatics Swimming YLDEFONSO, Teofilo PHI Men 200M Breaststroke Bronze
5889 1932 Los Angeles Athletics Athletics TORIBIO, Simeon Galvez PHI Men High Jump Bronze
5922 1932 Los Angeles Boxing Boxing VILLANUEVA, Jose PHI Men 50.8 - 54KG (Bantamweight) Bronze
6447 1936 Berlin Athletics Athletics WHITE, Miguel S. PHI Men 400M Hurdles Bronze
11005 1964 Tokyo Boxing Boxing VILLANUEVA, Anthony N. PHI Men 54 - 57KG (Featherweight) Silver
18513 1988 Seoul Boxing Boxing SERANTES, Leopoldo PHI Men - 48KG (Light-Flyweight) Bronze
20184 1992 Barcelona Boxing Boxing VELASCO, Roel PHI Men - 48KG (Light-Flyweight) Bronze
21927 1996 Atlanta Boxing Boxing VELASCO, Mansueto PHI Men - 48KG (Light-Flyweight) Silver 
split2 = summer.groupby(by = ["Country", "Gender"])

l2 = list(split2)
l2[:2]

[(('AFG', 'Men'),
         Year     City      Sport Discipline           Athlete Country Gender  \
  28965  2008  Beijing  Taekwondo  Taekwondo  NIKPAI, Rohullah     AFG    Men   
  30929  2012   London  Taekwondo  Taekwondo  NIKPAI, Rohullah     AFG    Men   
  
              Event   Medal  
  28965     - 58 KG  Bronze  
  30929  58 - 68 KG  Bronze  ),
 (('AHO', 'Men'),
         Year   City    Sport Discipline          Athlete Country Gender  \
  19323  1988  Seoul  Sailing    Sailing  BOERSMA, Jan D.     AHO    Men   
  
                       Event   Medal  
  19323  Board (Division Ii)  Silver  )]

len(l2)

236

l2[104]

l2[104][0]

l2[104][1]

split-apply-combine explained 

import pandas as pd

titanic = pd.read_csv("titanic.csv")

titanic_slice = titanic.iloc[:10, [2,3]]

titanic_slice
sex age
0 male 22.0
1 female 38.0
2 female 26.0
3 female 35.0
4 male 35.0
5 male NaN
6 male 54.0
7 male 2.0
8 female 27.0
9 female 14.0 
list(titanic_slice.groupby("sex"))[0][1]
sex age
1 female 38.0
2 female 26.0
3 female 35.0
8 female 27.0
9 female 14.0 
list(titanic_slice.groupby("sex"))[1][1]

titanic_slice.groupby("sex").mean()
age
sex
female 28.00
male 28.25 
titanic.groupby("sex").survived.sum()

sex
female    233
male      109
Name: survived, dtype: int64

titanic.groupby("sex")[["fare", "age"]].max()
fare age
sex
female 512.3292 63.0
male 512.3292 80.0 
new_df = titanic.groupby("sex").mean()

new_df
survived pclass age sibsp parch fare
sex
female 0.742038 2.159236 27.915709 0.694268 0.649682 44.479818
male 0.188908 2.389948 30.726645 0.429809 0.235702 25.523893 
%matplotlib inline
import matplotlib.pyplot as plt
plt.style.use("seaborn")

new_df.plot(kind = "bar", subplots = True, figsize = (8,15), fontsize = 13)
plt.show()

png

split-apply-combine applied 

import pandas as pd

summer = pd.read_csv("summer.csv")

summer.head()

summer.info()

medals_per_country = summer.groupby("Country").Medal.count().nlargest(n = 20)
medals_per_country

Country
USA    4585
URS    2049
GBR    1720
FRA    1396
GER    1305
ITA    1296
AUS    1189
HUN    1079
SWE    1044
NED     851
GDR     825
CHN     807
JPN     788
RUS     768
CAN     649
ROU     640
NOR     554
KOR     529
POL     511
DEN     507
Name: Medal, dtype: int64

%matplotlib inline
import matplotlib.pyplot as plt
plt.style.use("seaborn")

medals_per_country.plot(kind = "bar", figsize = (14, 8), fontsize = 14)
plt.xlabel("Country", fontsize = 13)
plt.ylabel("No. of Medals", fontsize = 13)
plt.title("Summer Olympic Games (Total Medals per Country)", fontsize = 16)
plt.show()

png

titanic = pd.read_csv("titanic.csv")

titanic.head()

titanic.info()

titanic.describe()

titanic.fare.mean()

32.204207968574636

titanic.groupby("pclass").fare.mean()

pclass
1    84.154687
2    20.662183
3    13.675550
Name: fare, dtype: float64

titanic.survived.sum()

342

titanic.survived.mean()

0.3838383838383838

titanic.groupby("sex").survived.mean()

sex
female    0.742038
male      0.188908
Name: survived, dtype: float64

titanic.groupby("pclass").survived.mean()

pclass
1    0.629630
2    0.472826
3    0.242363
Name: survived, dtype: float64

titanic["ad_chi"] = "adult"

titanic.loc[titanic.age < 18, "ad_chi"] = "child"

titanic.head(20)

titanic.ad_chi.value_counts()

adult    778
child    113
Name: ad_chi, dtype: int64

titanic.groupby("ad_chi").survived.mean()

ad_chi
adult    0.361183
child    0.539823
Name: survived, dtype: float64

titanic.groupby(["sex", "ad_chi"]).survived.count()

sex     ad_chi
female  adult     259
        child      55
male    adult     519
        child      58
Name: survived, dtype: int64

titanic.groupby(["sex", "ad_chi"]).survived.mean().sort_values(ascending = False)

sex     ad_chi
female  adult     0.752896
        child     0.690909
male    child     0.396552
        adult     0.165703
Name: survived, dtype: float64

w_and_c_first = titanic.groupby(["sex", "ad_chi"]).survived.mean().sort_values(ascending = False)

w_and_c_first.plot(kind = "bar", figsize = (14,8), fontsize = 14)
plt.xlabel("Groups", fontsize = 13)
plt.ylabel("Survival Rate", fontsize = 13)
plt.title("Titanic Survival Rate by Sex/Age-Groups", fontsize = 16)
plt.show()

png

Advanced Aggregation with agg() 

import pandas as pd

titanic = pd.read_csv("titanic.csv", usecols = ["survived", "pclass", "sex", "age", "fare"])

titanic.head()

titanic.groupby("sex").mean()
survived pclass age fare
sex
female 0.742038 2.159236 27.915709 44.479818
male 0.188908 2.389948 30.726645 25.523893 
titanic.groupby("sex").sum()
survived pclass age fare
sex
female 233 678 7286.00 13966.6628
male 109 1379 13919.17 14727.2865 
titanic.groupby("sex").agg(["mean", "sum", "min", "max"])
survived pclass age fare
mean sum min max mean sum min max mean sum min max mean sum min max
sex
female 0.742038 233 0 1 2.159236 678 1 3 27.915709 7286.00 0.75 63.0 44.479818 13966.6628 6.75 512.3292
male 0.188908 109 0 1 2.389948 1379 1 3 30.726645 13919.17 0.42 80.0 25.523893 14727.2865 0.00 512.3292 
titanic.groupby("sex").agg({"survived": ["sum", "mean"], "pclass": "mean", "age": ["mean", "median"], "fare": "max"})
survived pclass age fare
sum mean mean mean median max
sex
female 233 0.742038 2.159236 27.915709 27.0 512.3292
male 109 0.188908 2.389948 30.726645 29.0 512.3292

GroupBy Aggregation with Relabeling (new in Version 0.25) 

import pandas as pd

titanic = pd.read_csv("titanic.csv", usecols = ["survived", "pclass", "sex", "age", "fare"])

titanic.head()

titanic.groupby("sex").survived.mean()

sex
female    0.742038
male      0.188908
Name: survived, dtype: float64

titanic.groupby("sex").agg(survival_rate = ("survived", "mean"))
survival_rate
sex
female 0.742038
male 0.188908 
titanic.groupby("sex").agg({"survived": ["sum", "mean"], "age": ["mean"]})
survived age
sum mean mean
sex
female 233 0.742038 27.915709
male 109 0.188908 30.726645 
titanic.groupby("sex").agg(survived_total = ("survived", "sum"), 
                           survival_rate = ("survived", "mean"), mean_age = ("age", "mean"))
survived_total survival_rate mean_age
sex
female 233 0.742038 27.915709
male 109 0.188908 30.726645

Transformation with transform() 

import pandas as pd

titanic = pd.read_csv("titanic.csv")

titanic.head()

titanic.groupby(["sex", "pclass"]).survived.transform("mean")

0      0.135447
1      0.968085
2      0.500000
3      0.968085
4      0.135447
         ...   
886    0.157407
887    0.968085
888    0.500000
889    0.368852
890    0.135447
Name: survived, Length: 891, dtype: float64

titanic["group_surv_rate"] = titanic.groupby(["sex", "pclass"]).survived.transform("mean")

titanic.head()
survived pclass sex age sibsp parch fare embarked deck group_surv_rate
0 0 3 male 22.0 1 0 7.2500 S NaN 0.135447
1 1 1 female 38.0 1 0 71.2833 C C 0.968085
2 1 3 female 26.0 0 0 7.9250 S NaN 0.500000
3 1 1 female 35.0 1 0 53.1000 S C 0.968085
4 0 3 male 35.0 0 0 8.0500 S NaN 0.135447 
titanic["outliers"] = abs(titanic.survived-titanic.group_surv_rate)

titanic[titanic.outliers > 0.85]

Replacing NA Values by group-specific Values 

import pandas as pd

titanic = pd.read_csv("titanic.csv")

titanic.head(20)

titanic.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 891 entries, 0 to 890
Data columns (total 9 columns):
 #   Column    Non-Null Count  Dtype  
---  ------    --------------  -----  
 0   survived  891 non-null    int64  
 1   pclass    891 non-null    int64  
 2   sex       891 non-null    object 
 3   age       714 non-null    float64
 4   sibsp     891 non-null    int64  
 5   parch     891 non-null    int64  
 6   fare      891 non-null    float64
 7   embarked  889 non-null    object 
 8   deck      203 non-null    object 
dtypes: float64(2), int64(4), object(3)
memory usage: 62.8+ KB

mean_age = titanic.age.mean()
mean_age

29.69911764705882

titanic.age.fillna(mean_age)

0      22.000000
1      38.000000
2      26.000000
3      35.000000
4      35.000000
         ...    
886    27.000000
887    19.000000
888    29.699118
889    26.000000
890    32.000000
Name: age, Length: 891, dtype: float64

titanic.groupby(["sex", "pclass"]).age.mean()

sex     pclass
female  1         34.611765
        2         28.722973
        3         21.750000
male    1         41.281386
        2         30.740707
        3         26.507589
Name: age, dtype: float64

titanic["group_mean_age"] = titanic.groupby(["sex", "pclass"]).age.transform("mean")

titanic.head(20)
survived pclass sex age sibsp parch fare embarked deck group_mean_age
0 0 3 male 22.0 1 0 7.2500 S NaN 26.507589
1 1 1 female 38.0 1 0 71.2833 C C 34.611765
2 1 3 female 26.0 0 0 7.9250 S NaN 21.750000
3 1 1 female 35.0 1 0 53.1000 S C 34.611765
4 0 3 male 35.0 0 0 8.0500 S NaN 26.507589
5 0 3 male NaN 0 0 8.4583 Q NaN 26.507589
6 0 1 male 54.0 0 0 51.8625 S E 41.281386
7 0 3 male 2.0 3 1 21.0750 S NaN 26.507589
8 1 3 female 27.0 0 2 11.1333 S NaN 21.750000
9 1 2 female 14.0 1 0 30.0708 C NaN 28.722973
10 1 3 female 4.0 1 1 16.7000 S G 21.750000
11 1 1 female 58.0 0 0 26.5500 S C 34.611765
12 0 3 male 20.0 0 0 8.0500 S NaN 26.507589
13 0 3 male 39.0 1 5 31.2750 S NaN 26.507589
14 0 3 female 14.0 0 0 7.8542 S NaN 21.750000
15 1 2 female 55.0 0 0 16.0000 S NaN 28.722973
16 0 3 male 2.0 4 1 29.1250 Q NaN 26.507589
17 1 2 male NaN 0 0 13.0000 S NaN 30.740707
18 0 3 female 31.0 1 0 18.0000 S NaN 21.750000
19 1 3 female NaN 0 0 7.2250 C NaN 21.750000 
titanic.age.fillna(titanic.group_mean_age, inplace = True)

titanic.head(20)
survived pclass sex age sibsp parch fare embarked deck group_mean_age
0 0 3 male 22.000000 1 0 7.2500 S NaN 26.507589
1 1 1 female 38.000000 1 0 71.2833 C C 34.611765
2 1 3 female 26.000000 0 0 7.9250 S NaN 21.750000
3 1 1 female 35.000000 1 0 53.1000 S C 34.611765
4 0 3 male 35.000000 0 0 8.0500 S NaN 26.507589
5 0 3 male 26.507589 0 0 8.4583 Q NaN 26.507589
6 0 1 male 54.000000 0 0 51.8625 S E 41.281386
7 0 3 male 2.000000 3 1 21.0750 S NaN 26.507589
8 1 3 female 27.000000 0 2 11.1333 S NaN 21.750000
9 1 2 female 14.000000 1 0 30.0708 C NaN 28.722973
10 1 3 female 4.000000 1 1 16.7000 S G 21.750000
11 1 1 female 58.000000 0 0 26.5500 S C 34.611765
12 0 3 male 20.000000 0 0 8.0500 S NaN 26.507589
13 0 3 male 39.000000 1 5 31.2750 S NaN 26.507589
14 0 3 female 14.000000 0 0 7.8542 S NaN 21.750000
15 1 2 female 55.000000 0 0 16.0000 S NaN 28.722973
16 0 3 male 2.000000 4 1 29.1250 Q NaN 26.507589
17 1 2 male 30.740707 0 0 13.0000 S NaN 30.740707
18 0 3 female 31.000000 1 0 18.0000 S NaN 21.750000
19 1 3 female 21.750000 0 0 7.2250 C NaN 21.750000 
titanic.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 891 entries, 0 to 890
Data columns (total 10 columns):
 #   Column          Non-Null Count  Dtype  
---  ------          --------------  -----  
 0   survived        891 non-null    int64  
 1   pclass          891 non-null    int64  
 2   sex             891 non-null    object 
 3   age             891 non-null    float64
 4   sibsp           891 non-null    int64  
 5   parch           891 non-null    int64  
 6   fare            891 non-null    float64
 7   embarked        889 non-null    object 
 8   deck            203 non-null    object 
 9   group_mean_age  891 non-null    float64
dtypes: float64(3), int64(4), object(3)
memory usage: 69.7+ KB

Generalizing split-apply-combine with apply() 

import pandas as pd

titanic = pd.read_csv("titanic.csv", usecols = ["survived", "pclass", "sex", "age", "fare"])

titanic.head()

titanic.groupby("sex").mean()
survived pclass age fare
sex
female 0.742038 2.159236 27.915709 44.479818
male 0.188908 2.389948 30.726645 25.523893 
female_group = list(titanic.groupby("sex"))[0][1]
female_group
survived pclass sex age fare
1 1 1 female 38.0 71.2833
2 1 3 female 26.0 7.9250
3 1 1 female 35.0 53.1000
8 1 3 female 27.0 11.1333
9 1 2 female 14.0 30.0708
... ... ... ... ... ...
880 1 2 female 25.0 26.0000
882 0 3 female 22.0 10.5167
885 0 3 female 39.0 29.1250
887 1 1 female 19.0 30.0000
888 0 3 female NaN 23.4500

314 rows × 5 columns

female_group.mean().astype("float")

C:\Users\LENOVO\AppData\Local\Temp/ipykernel_19564/2434558135.py:1: FutureWarning: Dropping of nuisance columns in DataFrame reductions (with 'numeric_only=None') is deprecated; in a future version this will raise TypeError.  Select only valid columns before calling the reduction.
  female_group.mean().astype("float")





survived     0.742038
pclass       2.159236
age         27.915709
fare        44.479818
dtype: float64

def group_mean(group):
    return group.mean()

group_mean(female_group)

C:\Users\LENOVO\AppData\Local\Temp/ipykernel_19564/359042690.py:2: FutureWarning: Dropping of nuisance columns in DataFrame reductions (with 'numeric_only=None') is deprecated; in a future version this will raise TypeError.  Select only valid columns before calling the reduction.
  return group.mean()





survived     0.742038
pclass       2.159236
age         27.915709
fare        44.479818
dtype: float64

titanic.groupby("sex").apply(group_mean)

C:\Users\LENOVO\AppData\Local\Temp/ipykernel_19564/359042690.py:2: FutureWarning: Dropping of nuisance columns in DataFrame reductions (with 'numeric_only=None') is deprecated; in a future version this will raise TypeError.  Select only valid columns before calling the reduction.
  return group.mean()
survived pclass age fare
sex
female 0.742038 2.159236 27.915709 44.479818
male 0.188908 2.389948 30.726645 25.523893 
titanic.nlargest(5, "age")
survived pclass sex age fare
630 1 1 male 80.0 30.0000
851 0 3 male 74.0 7.7750
96 0 1 male 71.0 34.6542
493 0 1 male 71.0 49.5042
116 0 3 male 70.5 7.7500 
def five_oldest_surv(group):
    return group[group.survived == 1].nlargest(5, "age")

titanic.groupby("sex").apply(five_oldest_surv)

Hierarchical Indexing (MultiIndex) with Groupby 

import pandas as pd

titanic = pd.read_csv("titanic.csv", usecols = ["survived", "pclass", "sex", "age", "fare"])

titanic

summary = titanic.groupby(["sex", "pclass"]).mean()

summary
survived age fare
sex pclass
female 1 0.968085 34.611765 106.125798
2 0.921053 28.722973 21.970121
3 0.500000 21.750000 16.118810
male 1 0.368852 41.281386 67.226127
2 0.157407 30.740707 19.741782
3 0.135447 26.507589 12.661633 
summary.index

MultiIndex([('female', 1),
            ('female', 2),
            ('female', 3),
            (  'male', 1),
            (  'male', 2),
            (  'male', 3)],
           names=['sex', 'pclass'])

summary.loc[("female", 2), :]

survived     0.921053
age         28.722973
fare        21.970121
Name: (female, 2), dtype: float64

summary.loc[("female", 2), "age"]

summary.swaplevel().sort_index()
survived age fare
pclass sex
1 female 0.968085 34.611765 106.125798
male 0.368852 41.281386 67.226127
2 female 0.921053 28.722973 21.970121
male 0.157407 30.740707 19.741782
3 female 0.500000 21.750000 16.118810
male 0.135447 26.507589 12.661633 
summary.reset_index()
sex pclass survived age fare
0 female 1 0.968085 34.611765 106.125798
1 female 2 0.921053 28.722973 21.970121
2 female 3 0.500000 21.750000 16.118810
3 male 1 0.368852 41.281386 67.226127
4 male 2 0.157407 30.740707 19.741782
5 male 3 0.135447 26.507589 12.661633

stack() and unstack() 

import pandas as pd

summer = pd.read_csv("summer.csv")

summer.head()

medals_by_country = summer.groupby(["Country", "Medal"]).Medal.count()

medals_by_country

Country  Medal 
AFG      Bronze     2
AHO      Silver     1
ALG      Bronze     8
         Gold       5
         Silver     2
                   ..
ZIM      Gold      18
         Silver     4
ZZX      Bronze    10
         Gold      23
         Silver    15
Name: Medal, Length: 347, dtype: int64

medals_by_country.loc[("USA", "Gold")]

2235

medals_by_country.shape

(347,)

medals_by_country.unstack(level = -1)
Medal Bronze Gold Silver
Country
AFG 2.0 NaN NaN
AHO NaN NaN 1.0
ALG 8.0 5.0 2.0
ANZ 5.0 20.0 4.0
ARG 91.0 69.0 99.0
... ... ... ...
VIE NaN NaN 2.0
YUG 118.0 143.0 174.0
ZAM 1.0 NaN 1.0
ZIM 1.0 18.0 4.0
ZZX 10.0 23.0 15.0

147 rows × 3 columns

medals_by_country = medals_by_country.unstack(level = -1, fill_value= 0)

medals_by_country.head()

medals_by_country.shape

(147, 3)

medals_by_country = medals_by_country[["Gold", "Silver", "Bronze"]]

medals_by_country.sort_values(by = ["Gold", "Silver", "Bronze"], ascending = [False, False, False], inplace = True)

medals_by_country.head(10)
Medal Gold Silver Bronze
Country
USA 2235 1252 1098
URS 838 627 584
GBR 546 621 553
ITA 476 416 404
GER 452 378 475
HUN 412 316 351
FRA 408 491 497
SWE 349 367 328
GDR 329 271 225
AUS 312 405 472 
import matplotlib.pyplot as plt
plt.style.use("seaborn")

medals_by_country.head(10).plot(kind = "bar", figsize = (12,8), fontsize = 13)
plt.xlabel("Country", fontsize = 13)
plt.ylabel("Medals", fontsize = 13)
plt.title("Medals per Country", fontsize = 16)
plt.legend(fontsize = 15)
plt.show()

png

medals_by_country.stack().unstack()

Country  Medal 
USA      Gold      2235
         Silver    1252
         Bronze    1098
URS      Gold       838
         Silver     627
                   ... 
NIG      Silver       0
         Bronze       1
TOG      Gold         0
         Silver       0
         Bronze       1
Length: 441, dtype: int64