Numpy: Check array for string data type

Question

How can I determine if a Numpy array contains a string? The array a in

a = np.array('hi world')

has data type dtype('|S8'), where 8 refers to the number of characters in the string.

I don't see how regular expressions (such as re.match('\|S\d+', a.dtype)) would work here as the data type isn't simply '|S8'.

ACarter · Accepted Answer · 2024-09-05 12:17:04Z

25

a.dtype.char == 'S'

or

a.dtype.type is np.string_

for Python 3.x you'll need

a.dtype.type is np.str_

ACarter

5,7479 gold badges42 silver badges60 bronze badges

answered May 28, 2012 at 21:31

Fred Foo

365k80 gold badges765 silver badges852 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

I noticed there's also a.dtype.kind == 'S'. Sweet!

Just a note: seems that this needs to be a.dtype.type is np.str_ in Python 3.x

For Python 2.x and 3.x you could test against a.dtype.kind in {'U', 'S'} to catch strings and unicode.

What happens if you put strings in an object array with character 'O'?

a.dtype.type is np.string_ or a.dtype.type is np.str_ is a further alternative

|