pd.Series.str.removesuffix
pandas.Series.str.removesuffix(suffix)
Supported Arguments
argument |
datatypes |
suffix |
String |
>>> @bodo.jit
... def f(S):
... return S.str.removesuffix("b")
>>> S = pd.Series(["a", "ab", "abc", " abcd", "a bcd", "abcd", "xab"])
>>> f(S)
0 a
1 a
2 abc
3 abcd
4 a bcd
5 abcd
6 xa
dtype: string