Fix numpy vstack on generator expressions #12467
Conversation
|
oops, wrong branch. Let me rebase. |
|
shouldn't numpy avoid breaking changes? ;) |
e9b8301 to
b44107a
Compare
|
I agree but this is an easy workaround in the mean time. |
|
Note sure about the 0.20.1 milestone but probably safe to backport. |
|
ok but this change needs to be made in many more places, right? |
|
Hum you are rights, I have just fixed a specific module but there are many others in the travis report. I reported the issue upstream. Let's see what they think of it. I have to run now. Feel free to push the missing fix into that PR directly if you want. |
|
Related to numpy/numpy#12263 and the NEP implementing array function protocol. For SciPy we did a similar fix by switching to list comprehensions. @shoyer noted in the above issue the possibility of re-enabling the behavior for now, but eventually raising a |
|
I guess this is up for grabs if anyone wants to finish it as we should try to avoid the |
b44107a to
e52ca6d
Compare
|
I think I fixed them all (when running the tests locally with numpy master). |
| Miscellaneous | ||
| ............. | ||
|
|
||
| - |Fix| Make sure to avoid raising ``FutureWarning`` when calling |
There was a problem hiding this comment.
Seems that FutureWarning is only a proposal (not implemented) in numpy?
There was a problem hiding this comment.
It's pretty likely to be implemented. If not we can always update our changelog later.
There was a problem hiding this comment.
Could you give me some pointers about where the proposal/discussion about FutureWarning and VisibleDeprecationWarning in numpy is happening?
It'll effect the semantics behind sklearn.utils.testing.assert_no_warnings: d6972a9
There was a problem hiding this comment.
The FutureWarning PR has been merged into numpy master as well: https://github.com/numpy/numpy/pull/12280/files
|
looks good. haven't checked for completeness but cron job will do that for us. |
|
FYI, there's only a joblib issue in Travis cron job now :) |
* Workaround vstack issue with genxp * Use list comprehensions instead of genexps with np.vstack * Add changelog entry.
* Workaround vstack issue with genxp * Use list comprehensions instead of genexps with np.vstack * Add changelog entry.
* Workaround vstack issue with genxp * Use list comprehensions instead of genexps with np.vstack * Add changelog entry.
* Workaround vstack issue with genxp * Use list comprehensions instead of genexps with np.vstack * Add changelog entry.
This reverts commit 71607a8.
This reverts commit 71607a8.
* Workaround vstack issue with genxp * Use list comprehensions instead of genexps with np.vstack * Add changelog entry.
This workaround should fix the test failing with numpy master (CRON job on travis):
https://travis-ci.org/scikit-learn/scikit-learn/builds/446632038
407 >>> from sklearn.cluster import SpectralBiclustering 408 >>> import numpy as np 409 >>> X = np.array([[1, 1], [2, 1], [1, 0], 410 ... [4, 7], [3, 5], [3, 6]]) 411 >>> clustering = SpectralBiclustering(n_clusters=2, random_state=0).fit(X) UNEXPECTED EXCEPTION: ValueError('need at least one array to concatenate') Traceback (most recent call last): File "/home/travis/miniconda/envs/testenv/lib/python3.7/doctest.py", line 1329, in __run compileflags, 1), test.globs) File "<doctest sklearn.cluster.bicluster.SpectralBiclustering[3]>", line 1, in <module> File "/home/travis/build/scikit-learn/scikit-learn/sklearn/cluster/bicluster.py", line 124, in fit self._fit(X) File "/home/travis/build/scikit-learn/scikit-learn/sklearn/cluster/bicluster.py", line 508, in _fit for label in range(n_row_clusters) File "/home/travis/miniconda/envs/testenv/lib/python3.7/site-packages/numpy/core/overrides.py", line 151, in public_api implementation, public_api, relevant_args, args, kwargs) File "/home/travis/miniconda/envs/testenv/lib/python3.7/site-packages/numpy/core/overrides.py", line 96, in array_function_implementation_or_override return implementation(*args, **kwargs) File "/home/travis/miniconda/envs/testenv/lib/python3.7/site-packages/numpy/core/shape_base.py", line 260, in vstack return _nx.concatenate([atleast_2d(_m) for _m in tup], 0) File "/home/travis/miniconda/envs/testenv/lib/python3.7/site-packages/numpy/core/overrides.py", line 151, in public_api implementation, public_api, relevant_args, args, kwargs) File "/home/travis/miniconda/envs/testenv/lib/python3.7/site-packages/numpy/core/overrides.py", line 96, in array_function_implementation_or_override return implementation(*args, **kwargs) File "/home/travis/miniconda/envs/testenv/lib/python3.7/site-packages/numpy/core/multiarray.py", line 193, in concatenate return _multiarray_umath.concatenate(arrays, axis, out) ValueError: need at least one array to concatenate