Pandas sort column values with "0" values at end of column
Question:
I have a dataframe which looks like this
pd.DataFrame({'A': [5, 2, 0, 0, -3, -2, 1]}).sort_values('A')
Out[6]:
A
4 -3
5 -2
2 0
3 0
6 1
1 2
0 5
I would like to have "0" values at the end when sorting so my resulting dataframe looks like this.
A
4 -3
5 -2
6 1
1 2
0 5
2 0
3 0
Is there a simple (1 line of code) solution?
Answers:
Let’s try adding a new column and sort by two columns:
df.assign(dummy=df.A.eq(0)).sort_values(['dummy','A']).drop('dummy', axis=1)
Another option, not quite a one-liner, is mask and concat:
mask = df['A'].eq(0)
df = pd.concat([df[~mask].sort_values('A'), df[mask]])
Output:
A
4 -3
5 -2
6 1
1 2
0 5
2 0
3 0
First mask the zeros then use argsort on column A to get the indices that would sort the dataframe:
df.iloc[df['A'].replace(0, np.nan).to_numpy().argsort()]
A
4 -3
5 -2
6 1
1 2
0 5
2 0
3 0
Another option is to use key in sort_values()
df.sort_values(by='A', key=lambda col: col.replace(0, np.nan), na_position='last')
I have a dataframe which looks like this
pd.DataFrame({'A': [5, 2, 0, 0, -3, -2, 1]}).sort_values('A')
Out[6]:
A
4 -3
5 -2
2 0
3 0
6 1
1 2
0 5
I would like to have "0" values at the end when sorting so my resulting dataframe looks like this.
A
4 -3
5 -2
6 1
1 2
0 5
2 0
3 0
Is there a simple (1 line of code) solution?
Let’s try adding a new column and sort by two columns:
df.assign(dummy=df.A.eq(0)).sort_values(['dummy','A']).drop('dummy', axis=1)
Another option, not quite a one-liner, is mask and concat:
mask = df['A'].eq(0)
df = pd.concat([df[~mask].sort_values('A'), df[mask]])
Output:
A
4 -3
5 -2
6 1
1 2
0 5
2 0
3 0
First mask the zeros then use argsort on column A to get the indices that would sort the dataframe:
df.iloc[df['A'].replace(0, np.nan).to_numpy().argsort()]
A
4 -3
5 -2
6 1
1 2
0 5
2 0
3 0
Another option is to use key in sort_values()
df.sort_values(by='A', key=lambda col: col.replace(0, np.nan), na_position='last')