CODEWITHSUNDEEP
pythonpandasiteration
moys
moys

Reputation: 8033

Iretarting over column & creating min() column for each of the columns

I have dataframe as below

d = {'Name':['Alisa','Bobby','jodha','jack','raghu','Cathrine',
'Alisa','Bobby','kumar','Alisa','Alex','Cathrine'],
'Age':[26,24,23,22,23,24,26,24,22,23,24,24],
'Score':[85,63,55,74,31,77,85,63,42,62,89,77]}

df = pd.DataFrame(d,columns=['Name','Age','Score'])
Name    Age     Score
0   Alisa   26  85
1   Bobby   24  63
2   jodha   23  55
3   jack    22  74
4   raghu   23  31
5   Cathrine    24  77
6   Alisa   26  85
7   Bobby   24  63
8   kumar   22  42
9   Alisa   23  62
10  Alex    24  89
11  Cathrine    24  77

When i run the code below, it works fine & gets the output as shown. A new column for each for the column.

a=df.columns[1:]
df[a +'rat'] = df[a]/df[a].sum()
Name    Age     Score   Agerat  Scorerat
0   Alisa   26  85  0.091228    0.105853
1   Bobby   24  63  0.084211    0.078456
2   jodha   23  55  0.080702    0.068493
3   jack    22  74  0.077193    0.092154
4   raghu   23  31  0.080702    0.038605
5   Cathrine    24  77  0.084211    0.095890
6   Alisa   26  85  0.091228    0.105853
7   Bobby   24  63  0.084211    0.078456
8   kumar   22  42  0.077193    0.052304
9   Alisa   23  62  0.080702    0.077210
10  Alex    24  89  0.084211    0.110834
11  Cathrine    24  77  0.084211    0.095890

However, when i want to create a Min for each of the columns with the code a below, i get the error "KeyError: "None of [Index(['Agemin', 'Scoremin'], dtype='object')] are in the [columns]". I wanted it create that column!

df[a +'min'] = df[a].min()

so, how do we go about creating min(), max(), sum() etc columns for each of the columns without having to specify the names of each of the columns?

Upvotes: 2

Views: 57

Answers (2)

Allen Qin
Allen Qin

Reputation: 19947

If you would prefer to do it using your original code, you can do:

df[a +'min'] = df[a].groupby(by=np.zeros_like(df.index)).transform(min)

Upvotes: 2

ALollz
ALollz

Reputation: 59529

You can assign multiple scalar values from a Series using ** to pass the arguments. The index becomes the column name, with the value broadcast to all rows. For a Series add_suffix adds to the index, while later I use it to add to column names, which how that behaves for DataFrames.

import pandas as pd

df1 = df.select_dtypes('number')

df = df.assign(**df1.min().add_suffix('min'))
#        Name  Age  Score  Agemin  Scoremin
#0      Alisa   26     85      22        31
#1      Bobby   24     63      22        31
#2      jodha   23     55      22        31
...
#10      Alex   24     89      22        31
#11  Cathrine   24     77      22        31

Personally, I would concat the other result:

df = pd.concat([df, (df1/df1.sum()).add_suffix('rat')], axis=1)

#        Name  Age  Score  Agemin  Scoremin    Agerat  Scorerat
#0      Alisa   26     85      22        31  0.091228  0.105853
#1      Bobby   24     63      22        31  0.084211  0.078456
#2      jodha   23     55      22        31  0.080702  0.068493
#...
#10      Alex   24     89      22        31  0.084211  0.110834
#11  Cathrine   24     77      22        31  0.084211  0.095890

Upvotes: 2

Related Questions