Skip to content

Add methods from the builtin float types to the numpy floating point types #13375

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
2 of 5 tasks
eric-wieser opened this issue Apr 20, 2019 · 8 comments
Open
2 of 5 tasks

Add methods from the builtin float types to the numpy floating point types #13375

eric-wieser opened this issue Apr 20, 2019 · 8 comments
Labels
01 - Enhancement Project Possible project, may require specific skills and long commitment

Comments

@eric-wieser
Copy link
Member

eric-wieser commented Apr 20, 2019 •
edited by BvB93
Loading

Asking for methods that exist on builtin float types but not numpy types gives:

>>> set(dir(float)) - set(dir(np.float32))
{'__getformat__',
 '__getnewargs__',
 '__setformat__',
 '__trunc__',
 'as_integer_ratio',
 'fromhex',
 'hex',
 'is_integer'}

Filtering these to the ones that are not CPython implementation details, we have:

For all of these, the best approach is probably to port the CPython implementations, taking care to handle the extended precision of long doubles.

Note that we get these methods for free on np.float64 due to inheriting from float
@eric-wieser eric-wieser changed the title Add np.longdouble.fromhex Add methods from the builtin float types to the numpy floating point types Apr 20, 2019
@eric-wieser eric-wieser added the Project Possible project, may require specific skills and long commitment label Apr 20, 2019
@eric-wieser eric-wieser mentioned this issue May 11, 2019
Closed
@eric-wieser eric-wieser mentioned this issue Oct 31, 2019
Closed
@eric-wieser
Copy link
Member Author

eric-wieser commented Oct 1, 2020 •
edited
Loading

Cpython has int.is_integer now (python/cpython#6121), so we might want to add is_integer on all np.number subclasses.

kkaris added a commit to kkaris/depmap_analysis that referenced this issue Oct 16, 2020
@kkaris
Replace math.trunc -> np.trunc per github.com/numpy/numpy/issues/13375
73615cc
@kkaris kkaris mentioned this issue Oct 19, 2020
Merged
@eric-wieser eric-wieser mentioned this issue Nov 17, 2020
Closed
@mw66
Copy link

mw66 commented Nov 17, 2020

Need to fix for all the np float types: np.float32, numpy.float16 / np.single, np.half:

$ python
Python 3.6.7 |Anaconda, Inc.| (default, Oct 23 2018, 19:16:44) 
[GCC 7.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> a=np.array([1,2,3], dtype=np.float32)
>>> a[0].is_integer()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'numpy.float32' object has no attribute 'is_integer'

>>> a=np.array([1,2,3], dtype=np.half)
>>> type(a[0])
<class 'numpy.float16'>
>>> a[0].is_integer()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: 'numpy.float16' object has no attribute 'is_integer'

>>> np.__version__
'1.19.4'

@BvB93
Copy link
Member

BvB93 commented Jun 25, 2021 •
edited
Loading

I think we can also add __ceil__ and __floor__ to the list.

@BvB93 BvB93 mentioned this issue Jun 25, 2021
Merged
@BvB93 BvB93 mentioned this issue Aug 31, 2021
Merged
@BvB93
Copy link
Member

BvB93 commented Sep 2, 2021

With the merging of #19803 is_integer can be crossed off the list.

@ganesh-k13
Copy link
Member

I'll take a stab at hex and fromhex by referring to python/cpython@65fe25e

@ganesh-k13 ganesh-k13 mentioned this issue Oct 10, 2021
Open
11 tasks
@mdickinson
Copy link
Contributor

@ganesh-k13 You should be aware that there was a long-standing corner-case bug in the float.fromhex code, so apply some caution if you're copying the code directly from the original commit to the CPython source. Here's the relevant PR: python/cpython#27834

@ganesh-k13
Copy link
Member

Thanks for the heads up @mdickinson!

gmarkall added a commit to gmarkall/numba that referenced this issue Nov 30, 2022
@gmarkall
Fix numba#8628: Don't test math.trunc with non-float64 NumPy scalars
0720123 
See numpy/numpy#13375. `__trunc__` is not defined for NumPy scalars
other than `float64`, so we can't call `math.trunc(x)` in the CUDA
simulator for NumPy scalars `x` if they are not `float64` type.
@gmarkall gmarkall mentioned this issue Nov 30, 2022
Merged
@bjodah bjodah mentioned this issue Jan 19, 2024
Closed
@creiter64 creiter64 mentioned this issue Feb 5, 2024
Open
@ganesh-k13 ganesh-k13 mentioned this issue May 13, 2025
Open
8 tasks
@jorenham
Copy link
Member

Python 3.14 also adds a float.from_number, fyi.

@ganesh-k13 ganesh-k13 mentioned this issue May 28, 2025
Draft
4 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
01 - Enhancement Project Possible project, may require specific skills and long commitment
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
6 participants
@eric-wieser @mdickinson @mw66 @jorenham @ganesh-k13 @BvB93
pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy