FazBrowse GitHub Viewer
|
Trending
|
URL:
|
Home
Tools:
[Download Repo ZIP]
[View Raw Code]
[Original HTTPS Page]
numpy/numpy/dtypes.py at v2.5.0 · numpy/numpy · GitHub
numpy
/
numpy
Public
Uh oh!
There was an error while loading.
Please reload this page
.
Notifications
You must be signed in to change notification settings
Fork
12.9k
Star
33k
Code
Issues
1.9k
Pull requests
308
Actions
Projects
Wiki
Security and quality
0
Insights
Additional navigation options
Code
Issues
Pull requests
Actions
Projects
Wiki
Security and quality
Insights
Expand file tree
Breadcrumbs
numpy
/
numpy
/
dtypes.py
Copy path
More file actions
More file actions
Latest commit
History
History
History
86 lines (64 loc) · 2.84 KB
Breadcrumbs
numpy
/
numpy
/
dtypes.py
Copy path
File metadata and controls
86 lines (64 loc) · 2.84 KB
Raw
Copy raw file
Download raw file
Open symbols panel
Edit and raw actions
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
"""
This module is home to specific dtypes related functionality and their classes.
For more general information about dtypes, also see `numpy.dtype` and
:ref:`arrays.dtypes`.
Similar to the builtin ``types`` module, this submodule defines types (classes)
that are not widely used directly.
.. versionadded:: NumPy 1.25
The dtypes module is new in NumPy 1.25. Previously DType classes were
only accessible indirectly.
DType classes
-------------
The following are the classes of the corresponding NumPy dtype instances and
NumPy scalar types. The classes can be used in ``isinstance`` checks and can
also be instantiated or used directly. Direct use of these classes is not
typical, since their scalar counterparts (e.g. ``np.float64``) or strings
like ``"float64"`` can be used.
"""
# See doc/source/reference/routines.dtypes.rst for module-level docs
__all__
=
[
"register_dlpack_dtype"
]
def
register_dlpack_dtype
(
dlpack_key
,
dtype
,
/
):
"""
Register a NumPy dtype for a DLPack ``(code, bits)`` pair so that
`numpy.from_dlpack` can import it and ``ndarray.__dlpack__`` can export
it. Built-in dtype mappings take priority on import.
If you think a conflict is possible but unproblematic you may wrap this
into a try/except block as NumPy will raise an error if another DType
is already registered for the same (code, bits) pair.
While an error is raised, the export is registered even on error.
Registering an identical (not equal) dtype multiple times is allowed but
normally registration should happen at import time.
.. warning::
It is the responsibility of the registering user to ensure that the
mapping is valid.
.. note::
This function was added primarily for ``ml_dtypes`` and may be
replaced with a different mechanism in the future. It is intended
to be used by authors of user-defined dtypes and not end-users.
Parameters
----------
dlpack_key : tuple of int
``(dl_dtype_code, dl_bits)`` matching the DLPack ``DLDataType``,
lanes is assumed to be always 1, currently.
dtype : dtype
A NumPy dtype instance.
Raises
------
ValueError : If a conflicting registration was already done.
See Also
--------
numpy.from_dlpack
"""
# Deferred to avoid circular import.
from
numpy
.
_core
.
_multiarray_umath
import
_register_dlpack_dtype
return
_register_dlpack_dtype
(
dlpack_key
,
dtype
)
def
_add_dtype_helper
(
DType
,
alias
):
# Function to add DTypes a bit more conveniently without channeling them
# through `numpy._core._multiarray_umath` namespace or similar.
from
numpy
import
dtypes
setattr
(
dtypes
,
DType
.
__name__
,
DType
)
__all__
.
append
(
DType
.
__name__
)
if
alias
:
alias
=
alias
.
removeprefix
(
"numpy.dtypes."
)
setattr
(
dtypes
,
alias
,
DType
)
__all__
.
append
(
alias
)
Back
|
FazBrowse Home
|
New Git URL