How to fill a Python numpy chararray with spaces?

Question

For some reason I'm struggling to initialize a numpy.chararray with spaces.

This works:

char_array1 = np.chararray((3, 3))
char_array1[:] = 'a'
char_array1

Output:

chararray([['a', 'a', 'a'],
       ['a', 'a', 'a'],
       ['a', 'a', 'a']], 
      dtype='|S1')

This doesn't:

char_array2 = np.chararray((3, 3))
char_array2[:] = ' '
char_array2

Output:

chararray([['', '', ''],
       ['', '', ''],
       ['', '', '']], 
      dtype='|S1')

What is causing this? I can't see an option to strip the items or something.

From help(chararray): The chararray class exists for backwards compatibility with Numarray, it is not recommended for new development. Starting from numpy1.4, if one needs arrays of strings, it is recommended to use arrays of dtype object_, string_ or unicode_, and use the free function in the numpy.char module for fast vectorized string operations. — John La Rooy
– John La Rooy, Commented May 8, 2017 at 22:08

Bill · Accepted Answer · 2017-05-08 04:26:45Z

3

In fact char arrays do remove whitespace:

Versus a regular NumPy array of type str or unicode, this class adds the following functionality:

values automatically have whitespace removed from the end when indexed comparison operators automatically remove whitespace from the end when comparing values vectorized string operations are provided as methods (e.g. endswith) and infix operators (e.g. "+", "*", "%")

So the answer is use a regular array of type str or unicode:

char_array3 = np.empty((3, 3), dtype='str')
char_array3[:] = ' '
char_array3

Output:

array([[' ', ' ', ' '],
       [' ', ' ', ' '],
       [' ', ' ', ' ']], 
      dtype='|S1')

answered May 8, 2017 at 4:26

Bill

11.8k13 gold badges68 silver badges100 bronze badges

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

5 Comments

hpaulj Over a year ago

I've seen few, if any references to chararray on SO. dtype='S...' is common. And I've recommended the np.char functions on occasion.

Bill Over a year ago

@hpaulj I looked into the documentation on 'S' type protocol and in the scipy documentation it says "Note that str refers to either null terminated bytes or unicode strings depending on the Python version. In code targeting both Python 2 and 3 U or np.unicode_ should be used as a dtype for strings. See Note on string types.

Bill Over a year ago

I suppose it depends whether you want zero-terminated bytes (a or S), or use actual bytes (b), or actual strings (U).

hpaulj Over a year ago

I don't think there's a difference in how the data is stored, just a difference in how it is displayed. char_array2.tobytes() and char_array2.view(np.ndarray) show that the spaces are there in the databuffer.

Bill Over a year ago

I see. You're right. char_array2.nbytes and char_array3.nbytes are both 9. Not sure what the documentation means then by "For backward compatibility with Python 2 the 'S' and 'a' typestrings remain zero-terminated bytes".

Daniel · Accepted Answer · 2017-05-08 04:33:47Z

0

Just create your array with ndarray:

chararray = np.ndarray((3,3), dtype='S1')
chararray[:]=' '

gives:

array([[' ', ' ', ' '],
   [' ', ' ', ' '],
   [' ', ' ', ' ']], 
  dtype='|S1')

answered May 8, 2017 at 4:33

Daniel

42.9k4 gold badges57 silver badges82 bronze badges

Collectives™ on Stack Overflow

How to fill a Python numpy chararray with spaces?

2 Answers 2

5 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

5 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related