cudf.Series.from_masked_array

classmethod Series.from_masked_array(data, mask, null_count=None)

Create a Series with null-mask. This is equivalent to:

Series(data).set_mask(mask, null_count=null_count)

Parameters

data1D array-like

The values. Null values must not be skipped. They can appear as garbage values.

mask1D array-like

The null-mask. Valid values are marked as 1; otherwise 0. The mask bit given the data index idx is computed as:

(mask[idx // 8] >> (idx % 8)) & 1

null_countint, optional

The number of null values. If None, it is calculated automatically.

Returns

Series

Examples

>>> import cudf
>>> a = cudf.Series([1, 2, 3, None, 4, None])
>>> a
0       1
1       2
2       3
3    <NA>
4       4
5    <NA>
dtype: int64
>>> b = cudf.Series([10, 11, 12, 13, 14])
>>> cudf.Series.from_masked_array(data=b, mask=a._column.mask)
0      10
1      11
2      12
3    <NA>
4      14
dtype: int64