`numpy.dtypes:StringDType`

Summary

Create a StringDType instance.

StringDType can be used to store UTF-8 encoded variable-width strings in a NumPy array.

na_object : object, optional: Object used to represent missing data. If unset, the array will not use a missing data sentinel.
coerce : bool, optional: Whether or not items in an array-like passed to an array creation function that are neither a str or str subtype should be coerced to str. Defaults to True. If set to False, creating a StringDType array from an array-like containing entries that are not already strings will raise an error.

import numpy as np

✓

from numpy.dtypes import StringDType

✓

np.array(["hello", "world"], dtype=StringDType())

✗

arr = np.array(["hello", None, "world"],
               dtype=StringDType(na_object=None))

✓

arr

✗

arr[1] is None

✓

arr = np.array(["hello", np.nan, "world"],
               dtype=StringDType(na_object=np.nan))

✓

np.isnan(arr)

✗

np.array([1.2, object(), "hello world"],
         dtype=StringDType(coerce=False))

✓

np.array(["hello", "world"], dtype=StringDType(coerce=True))