Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2 - 3.4. Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
* Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753) * Numerous performance improvements in various areas, most notably indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL. * Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes: https://github.com/numpy/numpy/blob/maintenance/1.9.x/doc/release/1.9.0-note... Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Cheers, Julian Taylor
On Sun, Jun 8, 2014 at 2:34 PM, Julian Taylor <jtaylor.debian@googlemail.com
wrote:
Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes:
https://github.com/numpy/numpy/blob/maintenance/1.9.x/doc/release/1.9.0-note... Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Thanks for this, good job.
Chuck
On Sun, Jun 8, 2014 at 1:43 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Sun, Jun 8, 2014 at 2:34 PM, Julian Taylor < jtaylor.debian@googlemail.com> wrote:
Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes:
https://github.com/numpy/numpy/blob/maintenance/1.9.x/doc/release/1.9.0-note... Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Thanks for this, good job.
Chuck
This is the first release of numpy I have had any involvement with, and I am truly amazed at the amount of talent and dedication you guys put into it.
Big, big thank you to everyone!
Jaime
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
I take nothing ever happened to clean up the datetime64 timezone mess?
sigh.
-Chris
On Sun, Jun 8, 2014 at 1:34 PM, Julian Taylor <jtaylor.debian@googlemail.com
wrote:
Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes:
https://github.com/numpy/numpy/blob/maintenance/1.9.x/doc/release/1.9.0-note... Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Cheers, Julian Taylor
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Still working on it, it's unfortunately taking much more time than I'd anticipated.
Cheers, Sankarshan Mudkavi
On Jun 9, 2014, at 12:10 PM, Chris Barker chris.barker@noaa.gov wrote:
I take nothing ever happened to clean up the datetime64 timezone mess?
sigh.
-Chris
On Sun, Jun 8, 2014 at 1:34 PM, Julian Taylor jtaylor.debian@googlemail.com wrote: Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes: https://github.com/numpy/numpy/blob/maintenance/1.9.x/doc/release/1.9.0-note... Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Cheers, Julian Taylor
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
--
Christopher Barker, Ph.D. Oceanographer
Emergency Response Division NOAA/NOS/OR&R (206) 526-6959 voice 7600 Sand Point Way NE (206) 526-6329 fax Seattle, WA 98115 (206) 526-6317 main reception
Chris.Barker@noaa.gov _______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On 6/8/2014 1:34 PM, Julian Taylor wrote:
Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes: https://github.com/numpy/numpy/blob/maintenance/1.9.x/doc/release/1.9.0-note... Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Cheers, Julian Taylor
Hello,
I tested numpy-MKL-1.9.0b1 (msvc9, Intel MKL build) on win-amd64-py2.7 against a few other packages that were built against numpy-MKL-1.8.x.
While numpy and scipy pass all tests, some other packages (matplotlib, statsmodels, skimage, pandas, pytables, sklearn...) show a few new test failures (compared to testing with numpy-MKL-1.8.1). Many test errors are of kind:
ValueError: shape mismatch: value array of shape (24,) could not be broadcast to indexing result of shape (8,3)
I have attached a list of failing tests. The full test results are at http://www.lfd.uci.edu/~gohlke/pythonlibs/tests/20140609-win-amd64-py2.7-numpy-1.9.0b1/ (compare to http://www.lfd.uci.edu/~gohlke/pythonlibs/tests/20140609-win-amd64-py2.7/)
I have not investigated any further...
Christoph
On Mon, Jun 9, 2014 at 5:21 PM, Christoph Gohlke cgohlke@uci.edu wrote:
On 6/8/2014 1:34 PM, Julian Taylor wrote:
Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes: https://github.com/numpy/numpy/blob/maintenance/1.9.x/ doc/release/1.9.0-notes.rst Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Cheers, Julian Taylor
Hello,
I tested numpy-MKL-1.9.0b1 (msvc9, Intel MKL build) on win-amd64-py2.7 against a few other packages that were built against numpy-MKL-1.8.x.
While numpy and scipy pass all tests, some other packages (matplotlib, statsmodels, skimage, pandas, pytables, sklearn...) show a few new test failures (compared to testing with numpy-MKL-1.8.1). Many test errors are of kind:
ValueError: shape mismatch: value array of shape (24,) could not be
broadcast to indexing result of shape (8,3)
I have attached a list of failing tests. The full test results are at < http://www.lfd.uci.edu/~gohlke/pythonlibs/tests/20140609-win-amd64-py2.7- numpy-1.9.0b1/> (compare to http://www.lfd.uci.edu/~ gohlke/pythonlibs/tests/20140609-win-amd64-py2.7/)
I have not investigated any further...
One of the matplotlib failures, and I suspect the others, comes from the assignment
found_index[refi_triangles[ancestor_mask, :] ] = np.repeat(ancestors[ancestor_mask], 3)
This fails with the error
ValueError: shape mismatch: value array of shape (13824,) could not be broadcast to indexing result of shape (4608,3)
I confess I find the construction odd, but the error probably results from stricter indexing rules. Indeed, (13824,) does not broadcast to (4608,3). Apart from considerations of backward compatibility, should it?
Chuck
On Mon, Jun 9, 2014 at 6:10 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 5:21 PM, Christoph Gohlke cgohlke@uci.edu wrote:
On 6/8/2014 1:34 PM, Julian Taylor wrote:
Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes: https://github.com/numpy/numpy/blob/maintenance/1.9.x/ doc/release/1.9.0-notes.rst Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Cheers, Julian Taylor
Hello,
I tested numpy-MKL-1.9.0b1 (msvc9, Intel MKL build) on win-amd64-py2.7 against a few other packages that were built against numpy-MKL-1.8.x.
While numpy and scipy pass all tests, some other packages (matplotlib, statsmodels, skimage, pandas, pytables, sklearn...) show a few new test failures (compared to testing with numpy-MKL-1.8.1). Many test errors are of kind:
ValueError: shape mismatch: value array of shape (24,) could not be
broadcast to indexing result of shape (8,3)
I have attached a list of failing tests. The full test results are at < http://www.lfd.uci.edu/~gohlke/pythonlibs/tests/20140609-win-amd64-py2.7- numpy-1.9.0b1/> (compare to http://www.lfd.uci.edu/~ gohlke/pythonlibs/tests/20140609-win-amd64-py2.7/)
I have not investigated any further...
One of the matplotlib failures, and I suspect the others, comes from the assignment
found_index[refi_triangles[ancestor_mask, :] ] = np.repeat(ancestors[ancestor_mask], 3)
This fails with the error
ValueError: shape mismatch: value array of shape (13824,)
could not be broadcast to indexing result of shape (4608,3)
I confess I find the construction odd, but the error probably results from stricter indexing rules. Indeed, (13824,) does not broadcast to (4608,3). Apart from considerations of backward compatibility, should it?
Other errors are of the type:
TypeError: NumPy boolean array indexing assignment requires a 0 or 1-dimensionalinput, input has 2 dimensions
This one looks to arise is from stricter rules for boolean indexing.
Chuck
On Mo, 2014-06-09 at 18:21 -0600, Charles R Harris wrote:
On Mon, Jun 9, 2014 at 6:10 PM, Charles R Harris charlesr.harris@gmail.com wrote:
<snip>
Other errors are of the type: TypeError: NumPy boolean array indexing assignment requires a 0 or 1-dimensionalinput, input has 2 dimensions This one looks to arise is from stricter rules for boolean indexing.
Hmmmm, I may have removed an "if size of the boolean index matches the size of the output array ignore the shape (dimensions)" special cased, I guess we can put that in again.
The other error looks a bit different because of the nonzero logic, but probably is the same, i.e. also boolean indexing. The last one is the change that `arr[[1,2,3,4]] = [1,2]` does not work anymore. A workaround (maybe also for the rest possibly) is `arr.flat[[1,2,3,4]] = [1,2]`, but I guess workarounds are not an option with matplotlib, so have to think about it.
- Sebastian
Chuck
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On Di, 2014-06-10 at 10:43 +0200, Sebastian Berg wrote:
On Mo, 2014-06-09 at 18:21 -0600, Charles R Harris wrote:
On Mon, Jun 9, 2014 at 6:10 PM, Charles R Harris charlesr.harris@gmail.com wrote:
<snip> > > > > > Other errors are of the type: > TypeError: NumPy boolean array indexing assignment requires a 0 or 1-dimensionalinput, input has 2 dimensions > This one looks to arise is from stricter rules for boolean indexing. >
Hmmmm, I may have removed an "if size of the boolean index matches the size of the output array ignore the shape (dimensions)" special cased, I guess we can put that in again.
The other error looks a bit different because of the nonzero logic, but probably is the same, i.e. also boolean indexing. The last one is the change that `arr[[1,2,3,4]] = [1,2]` does not work anymore. A workaround (maybe also for the rest possibly) is `arr.flat[[1,2,3,4]] = [1,2]`, but I guess workarounds are not an option with matplotlib, so have to think about it.
Correction: I think all (indexing) errors are the second case since the boolean special case is not taken.
- Sebastian
Chuck
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On 10 Jun 2014 09:44, "Sebastian Berg" sebastian@sipsolutions.net wrote:
The other error looks a bit different because of the nonzero logic, but probably is the same, i.e. also boolean indexing. The last one is the change that `arr[[1,2,3,4]] = [1,2]` does not work anymore. A workaround (maybe also for the rest possibly) is `arr.flat[[1,2,3,4]] = [1,2]`, but I guess workarounds are not an option with matplotlib, so have to think about it.
If the beta is breaking code then let's put the change off to 1.10 or so and raise a deprecation warning in 1.9.
-n
On Di, 2014-06-10 at 10:50 +0100, Nathaniel Smith wrote:
On 10 Jun 2014 09:44, "Sebastian Berg" sebastian@sipsolutions.net wrote:
The other error looks a bit different because of the nonzero logic,
but
probably is the same, i.e. also boolean indexing. The last one is
the
change that `arr[[1,2,3,4]] = [1,2]` does not work anymore. A
workaround
(maybe also for the rest possibly) is `arr.flat[[1,2,3,4]] = [1,2]`,
but
I guess workarounds are not an option with matplotlib, so have to
think
about it.
If the beta is breaking code then let's put the change off to 1.10 or so and raise a deprecation warning in 1.9.
Yes, unfortunately it is a bit more complicating than that.
-n
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On 10 Jun 2014 11:15, "Sebastian Berg" sebastian@sipsolutions.net wrote:
On Di, 2014-06-10 at 10:50 +0100, Nathaniel Smith wrote:
On 10 Jun 2014 09:44, "Sebastian Berg" sebastian@sipsolutions.net wrote:
The other error looks a bit different because of the nonzero logic,
but
probably is the same, i.e. also boolean indexing. The last one is
the
change that `arr[[1,2,3,4]] = [1,2]` does not work anymore. A
workaround
(maybe also for the rest possibly) is `arr.flat[[1,2,3,4]] = [1,2]`,
but
I guess workarounds are not an option with matplotlib, so have to
think
about it.
If the beta is breaking code then let's put the change off to 1.10 or so and raise a deprecation warning in 1.9.
Yes, unfortunately it is a bit more complicating than that.
Is it impossible to emulate the old arr[[1, 2, 3, 4]] = [1, 2] behavior for some reason? Or what do you mean? (I'm not suggesting we literally go back to the 1.8 indexing code.)
-n
On Di, 2014-06-10 at 11:24 +0100, Nathaniel Smith wrote:
On 10 Jun 2014 11:15, "Sebastian Berg" sebastian@sipsolutions.net wrote:
On Di, 2014-06-10 at 10:50 +0100, Nathaniel Smith wrote:
On 10 Jun 2014 09:44, "Sebastian Berg"
wrote:
The other error looks a bit different because of the nonzero
logic,
but
probably is the same, i.e. also boolean indexing. The last one
is
the
change that `arr[[1,2,3,4]] = [1,2]` does not work anymore. A
workaround
(maybe also for the rest possibly) is `arr.flat[[1,2,3,4]] =
[1,2]`,
but
I guess workarounds are not an option with matplotlib, so have
to
think
about it.
If the beta is breaking code then let's put the change off to 1.10
or
so and raise a deprecation warning in 1.9.
Yes, unfortunately it is a bit more complicating than that.
Is it impossible to emulate the old arr[[1, 2, 3, 4]] = [1, 2] behavior for some reason? Or what do you mean? (I'm not suggesting we literally go back to the 1.8 indexing code.)
Yeah, just have to check carefully. Maybe easiest is to just try/except it in C-code, throw a warning, and (if applicable) try calling the `arr.flat[...] = ...` code (which still exists).
- Sebastian
-n
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
Charles R Harris charlesr.harris@gmail.com wrote:
I confess I find the construction odd, but the error probably results from stricter indexing rules. Indeed, (13824,) does not broadcast to (4608,3). Apart from considerations of backward compatibility, should it?
Probably not. (But breaking matplotlib is bad.)
The one pandas test failure that is valid: ERROR: test_interp_regression (pandas.tests.test_generic.TestSeries)
has been fixed in pandas master / 0.14.1 (prob releasing in 1 month).
(the other test failures are for clipboard / network issues)
On Mon, Jun 9, 2014 at 7:21 PM, Christoph Gohlke cgohlke@uci.edu wrote:
On 6/8/2014 1:34 PM, Julian Taylor wrote:
Hello,
I'm happy to announce the fist beta release of Numpy 1.9.0. 1.9.0 will be a new feature release supporting Python 2.6 - 2.7 and 3.2
- 3.4.
Due to low demand windows binaries for the beta are only available for Python 2.7, 3.3 and 3.4. Please try it and report any issues to the numpy-discussion mailing list or on github.
The 1.9 release will consists of mainly of many small improvements and bugfixes. The highlights are:
- Addition of __numpy_ufunc__ to allow overriding ufuncs in ndarray
subclasses. Please note that there are still some known issues with this mechanism which we hope to resolve before the final release (e.g. #4753)
- Numerous performance improvements in various areas, most notably
indexing and operations on small arrays are significantly faster. Indexing operations now also release the GIL.
- Addition of nanmedian and nanpercentile rounds out the nanfunction set.
The changes involve a lot of small changes that might affect some applications, please read the release notes for the full details on all changes: https://github.com/numpy/numpy/blob/maintenance/1.9.x/ doc/release/1.9.0-notes.rst Please also take special note of the future changes section which will apply to the following release 1.10.0 and make sure to check if your applications would be affected by them.
Source tarballs, windows installers and release notes can be found at https://sourceforge.net/projects/numpy/files/NumPy/1.9.0b1
Cheers, Julian Taylor
Hello,
I tested numpy-MKL-1.9.0b1 (msvc9, Intel MKL build) on win-amd64-py2.7 against a few other packages that were built against numpy-MKL-1.8.x.
While numpy and scipy pass all tests, some other packages (matplotlib, statsmodels, skimage, pandas, pytables, sklearn...) show a few new test failures (compared to testing with numpy-MKL-1.8.1). Many test errors are of kind:
ValueError: shape mismatch: value array of shape (24,) could not be
broadcast to indexing result of shape (8,3)
I have attached a list of failing tests. The full test results are at < http://www.lfd.uci.edu/~gohlke/pythonlibs/tests/20140609-win-amd64-py2.7- numpy-1.9.0b1/> (compare to http://www.lfd.uci.edu/~ gohlke/pythonlibs/tests/20140609-win-amd64-py2.7/)
I have not investigated any further...
Christoph
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On Mon, Jun 9, 2014 at 6:21 PM, Jeff Reback jeffreback@gmail.com wrote:
The one pandas test failure that is valid: ERROR: test_interp_regression (pandas.tests.test_generic.TestSeries)
has been fixed in pandas master / 0.14.1 (prob releasing in 1 month).
(the other test failures are for clipboard / network issues)
Thanks for the feedback.
<snip>
Chuck
On Mon, Jun 9, 2014 at 6:23 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:21 PM, Jeff Reback jeffreback@gmail.com wrote:
The one pandas test failure that is valid: ERROR: test_interp_regression (pandas.tests.test_generic.TestSeries)
has been fixed in pandas master / 0.14.1 (prob releasing in 1 month).
(the other test failures are for clipboard / network issues)
Thanks for the feedback.
<snip>
Looks to me like a fair number of the other test failures are actually bugs in the tests, or could be argued to be so.
Chuck
On Mon, Jun 9, 2014 at 6:47 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:23 PM, Charles R Harris < charlesr.harris@gmail.com> wrote:
On Mon, Jun 9, 2014 at 6:21 PM, Jeff Reback jeffreback@gmail.com wrote:
The one pandas test failure that is valid: ERROR: test_interp_regression (pandas.tests.test_generic.TestSeries)
has been fixed in pandas master / 0.14.1 (prob releasing in 1 month).
(the other test failures are for clipboard / network issues)
Thanks for the feedback.
<snip>
Looks to me like a fair number of the other test failures are actually bugs in the tests, or could be argued to be so.
Or rather, actual bugs, but not in numpy. I can see this will take a while to settle ;)
Chuck
On Mon, Jun 9, 2014 at 8:49 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:47 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:23 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:21 PM, Jeff Reback jeffreback@gmail.com wrote:
The one pandas test failure that is valid: ERROR: test_interp_regression (pandas.tests.test_generic.TestSeries)
has been fixed in pandas master / 0.14.1 (prob releasing in 1 month).
(the other test failures are for clipboard / network issues)
Thanks for the feedback.
<snip>
Looks to me like a fair number of the other test failures are actually bugs in the tests, or could be argued to be so.
Or rather, actual bugs, but not in numpy. I can see this will take a while to settle ;)
not really a bug on our side given the old numpy behavior I was surprised about the failure because it still produces the correct result.
params[nuis_param_index] = nuisance_params ValueError: shape mismatch: value array of shape (2,) could not be broadcast to indexing result of shape (0,)
As far as I have figured out based on Christoph's test failures
`nuis_param_index` is array([], dtype=int32) and because numpy didn't assign anything, `nuisance_params` was picked essentially arbitrary along this code path
x = np.arange(5.) we_dont_care_what_this_is = np.random.randn(10) x[[]] = we_dont_care_what_this_is x
array([ 0., 1., 2., 3., 4.])
Does x[[]] = [] work with numpy 1.9?
Josef
Chuck
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On Mon, Jun 9, 2014 at 11:10 PM, josef.pktd@gmail.com wrote:
On Mon, Jun 9, 2014 at 8:49 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:47 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:23 PM, Charles R Harris charlesr.harris@gmail.com wrote:
On Mon, Jun 9, 2014 at 6:21 PM, Jeff Reback jeffreback@gmail.com wrote:
The one pandas test failure that is valid: ERROR: test_interp_regression (pandas.tests.test_generic.TestSeries)
has been fixed in pandas master / 0.14.1 (prob releasing in 1 month).
(the other test failures are for clipboard / network issues)
Thanks for the feedback.
<snip>
Looks to me like a fair number of the other test failures are actually bugs in the tests, or could be argued to be so.
Or rather, actual bugs, but not in numpy. I can see this will take a while to settle ;)
not really a bug on our side given the old numpy behavior
forgot to define: our side = statsmodels
I was surprised about the failure because it still produces the correct result.
params[nuis_param_index] = nuisance_params ValueError: shape mismatch: value array of shape (2,) could not be broadcast to indexing result of shape (0,)
As far as I have figured out based on Christoph's test failures
`nuis_param_index` is array([], dtype=int32) and because numpy didn't assign anything, `nuisance_params` was picked essentially arbitrary along this code path
x = np.arange(5.) we_dont_care_what_this_is = np.random.randn(10) x[[]] = we_dont_care_what_this_is x
array([ 0., 1., 2., 3., 4.])
Does x[[]] = [] work with numpy 1.9?
Josef
Chuck
NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion
On Mo, 2014-06-09 at 16:21 -0700, Christoph Gohlke wrote:
On 6/8/2014 1:34 PM, Julian Taylor wrote:
Hello,
<snip>
Hello,
I tested numpy-MKL-1.9.0b1 (msvc9, Intel MKL build) on win-amd64-py2.7 against a few other packages that were built against numpy-MKL-1.8.x.
While numpy and scipy pass all tests, some other packages (matplotlib, statsmodels, skimage, pandas, pytables, sklearn...) show a few new test failures (compared to testing with numpy-MKL-1.8.1). Many test errors are of kind:
ValueError: shape mismatch: value array of shape (24,) could not be
broadcast to indexing result of shape (8,3)
I have attached a list of failing tests. The full test results are at http://www.lfd.uci.edu/~gohlke/pythonlibs/tests/20140609-win-amd64-py2.7-numpy-1.9.0b1/ (compare to http://www.lfd.uci.edu/~gohlke/pythonlibs/tests/20140609-win-amd64-py2.7/)
I have not investigated any further...
I have put up a sketch for the indexing fix at https://github.com/numpy/numpy/pull/4804 not sure when I will have time to finish it off, so if someone has, go ahead ;). It should fix all those indexing errors (though numpy has harmless test failures with it currently). Deprecationwarnings (and not there yet FutureWarnings for the error type change) should be given then.
- Sebastian
Christoph _______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion