bigframes.pandas.Series.mask

Series.mask(cond, other=None) → Series

Replace values where the condition is True.

Examples:

>>> s = bpd.Series([10, 11, 12, 13, 14])
>>> s
0    10
1    11
2    12
3    13
4    14
dtype: Int64

You can mask the values in the Series based on a condition. The values matching the condition would be masked. The condition can be provided in formm of a Series.

>>> s.mask(s % 2 == 0)
0    <NA>
1      11
2    <NA>
3      13
4    <NA>
dtype: Int64

You can specify a custom mask value.

>>> s.mask(s % 2 == 0, -1)
0    -1
1    11
2    -1
3    13
4    -1
dtype: Int64
>>> s.mask(s % 2 == 0, 100*s)
0    1000
1      11
2    1200
3      13
4    1400
dtype: Int64

You can also use a remote function to evaluate the mask condition. This is useful in situation such as the following, where the mask condition is evaluated based on a complicated business logic which cannot be expressed in form of a Series.

>>> @bpd.remote_function(reuse=False, cloud_function_service_account="default")
... def should_mask(name: str) -> bool:
...     hash = 0
...     for char_ in name:
...         hash += ord(char_)
...     return hash % 2 == 0

>>> s = bpd.Series(["Alice", "Bob", "Caroline"])
>>> s
0       Alice
1         Bob
2    Caroline
dtype: string
>>> s.mask(should_mask)
0        <NA>
1         Bob
2    Caroline
dtype: string
>>> s.mask(should_mask, "REDACTED")
0    REDACTED
1         Bob
2    Caroline
dtype: string

Simple vectorized (i.e. they only perform operations supported on a Series) lambdas or python functions can be used directly.

>>> nums = bpd.Series([1, 2, 3, 4], name="nums")
>>> nums
0    1
1    2
2    3
3    4
Name: nums, dtype: Int64
>>> nums.mask(lambda x: (x+1) % 2 == 1)
0        1
1     <NA>
2        3
3     <NA>
Name: nums, dtype: Int64

>>> def is_odd(num):
...     return num % 2 == 1
>>> nums.mask(is_odd)
0     <NA>
1        2
2     <NA>
3        4
Name: nums, dtype: Int64

Parameters:

• cond (bool Series/DataFrame, array-like, or callable) – Where cond is False, keep the original value. Where True, replace with corresponding value from other. If cond is callable, it is computed on the Series/DataFrame and should return boolean Series/DataFrame or array. The callable must not change input Series/DataFrame (though pandas doesn’t check it).

• other (scalar, Series/DataFrame, or callable) – Entries where cond is True are replaced with corresponding value from other. If other is callable, it is computed on the Series/DataFrame and should return scalar or Series/DataFrame. The callable must not change input Series/DataFrame (though pandas doesn’t check it). If not specified, entries will be filled with the corresponding NULL value (np.nan for numpy dtypes, pd.NA for extension dtypes).

Returns:

Series after the replacement.

Return type:

bigframes.pandas.Series