For each element in `a`, return a list of the words in the string, using `sep` as the delimiter string. Calls :meth:`str.split` element-wise. Parameters ---------- a : array-like, with ``StringDType``, ``bytes_``, or ``str_`` dtype sep : str or unicode, optional
(a, sep=None, maxsplit=None)
| 1398 | |
| 1399 | @array_function_dispatch(_split_dispatcher) |
| 1400 | def _split(a, sep=None, maxsplit=None): |
| 1401 | """ |
| 1402 | For each element in `a`, return a list of the words in the |
| 1403 | string, using `sep` as the delimiter string. |
| 1404 | |
| 1405 | Calls :meth:`str.split` element-wise. |
| 1406 | |
| 1407 | Parameters |
| 1408 | ---------- |
| 1409 | a : array-like, with ``StringDType``, ``bytes_``, or ``str_`` dtype |
| 1410 | |
| 1411 | sep : str or unicode, optional |
| 1412 | If `sep` is not specified or None, any whitespace string is a |
| 1413 | separator. |
| 1414 | |
| 1415 | maxsplit : int, optional |
| 1416 | If `maxsplit` is given, at most `maxsplit` splits are done. |
| 1417 | |
| 1418 | Returns |
| 1419 | ------- |
| 1420 | out : ndarray |
| 1421 | Array of list objects |
| 1422 | |
| 1423 | Examples |
| 1424 | -------- |
| 1425 | >>> import numpy as np |
| 1426 | >>> x = np.array("Numpy is nice!") |
| 1427 | >>> np.strings.split(x, " ") # doctest: +SKIP |
| 1428 | array(list(['Numpy', 'is', 'nice!']), dtype=object) # doctest: +SKIP |
| 1429 | |
| 1430 | >>> np.strings.split(x, " ", 1) # doctest: +SKIP |
| 1431 | array(list(['Numpy', 'is nice!']), dtype=object) # doctest: +SKIP |
| 1432 | |
| 1433 | See Also |
| 1434 | -------- |
| 1435 | str.split, rsplit |
| 1436 | |
| 1437 | """ |
| 1438 | # This will return an array of lists of different sizes, so we |
| 1439 | # leave it as an object array |
| 1440 | return _vec_string( |
| 1441 | a, np.object_, 'split', [sep] + _clean_args(maxsplit)) |
| 1442 | |
| 1443 | |
| 1444 | @array_function_dispatch(_split_dispatcher) |
nothing calls this directly
no test coverage detected
searching dependent graphs…