If I have this value.counts() dataframe (already ascending) :

A 20
B 15
C 15
D 10
E 10
F 10
G  8
H  5
I  5

Then I want to get a first 70% for example, then what I get is

A
B
C
D
E

Do pandas have function for that ? I have tried with groupby but it does not work. Or should I code manually like with for loop or something ?

thanks

Find biggest coverage in percent with pandas from a dataframe. Is there any shortcut function or should code manually ?

CodePudding user response：

Assuming X, Y the column names, you can compare the cumsum to be lower or equal (le) to 70, and slice with boolean indexing:

df.loc[df['Y'].cumsum().le(70), 'X']

Alternative by position (first and second column):

df.loc[df.iloc[:, 1].cumsum().le(70), df.columns[0]]

output:

0    A
1    B
2    C
3    D
4    E
Name: X, dtype: object

Used input:

   X   Y
0  A  20
1  B  15
2  C  15
3  D  10
4  E  10
5  F  10
6  G   8
7  H   5
8  I   5
9  G   2