[Python] Using split on a column for subset of rows

I'm trying to selectively parse rows based on a condition:

df = pd.DataFrame({"a":["z|v", "b|3", "z|x", "dfe|5", "xs|2"]})
df['x']=""
df['y']=""

mask = df['a'].str.startswith('z')
print(mask)

df.loc[mask, ['x', 'y']] = df.loc[mask, 'a'].str.split('|', n=1, expand=True)
print(df)


Outputs:

a x y
0 z|v NaN NaN
1 b|3
2 z|x NaN NaN
3 dfe|5
4 xs|2


its returning NaNs.

if I just apply:

df[['x', 'y']] = df['a'].str.split('|', n=1, expand=True)


it works for all rows:

a x y
0 z|v z v
1 b|3 b 3
2 z|x z x
3 dfe|5 dfe 5
4 xs|2 xs 2


Any help appreciated.

Continue reading...