Raspbian automatic forward porter [Mon, 21 Sep 2020 08:39:38 +0000 (09:39 +0100)]
Merge version 0.11.1-3+rpi1 and 0.12.0-2 to produce 0.12.0-2+rpi1
Rebecca N. Palmer [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Merge statsmodels (0.12.0-2) import into refs/heads/workingbranch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Ignore test failures on arm64, warn on armel
arm64 - TestZeroInflatedModel_probit convergence issues
armel - 21 crashes, 2 wrong answers (ignored in d/rules)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/968210
Forwarded: no (current form does not work on non-multiarch systems)
Gbp-Pq: Name 968210_ignore_tests.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Remove code from Stack Overflow and wasteful loop
Stack Overflow content is CC-BY-SA licensed,
which this package is not supposed to be. This snippet may be
too small to be copyrightable, but removing it to be safe.
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name remove_ccbysa_snippet.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Allow tests to fail if multiprocessing is not available
This is currently the case on hurd-i386
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_no_multiprocessing.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
DynamicFactor test fail on ppc64el
Warn on use and xfail test
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_dynamicfactor_ppc64el.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Markov* broken on armhf
Warn on use, and xfail tests
I suspect https://github.com/statsmodels/statsmodels/pull/5826
but that improved it enough on amd64 that
reverting it isn't a good solution.
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_regimeswitching_armhf.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Fix typos
Origin: lintian
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name spelling.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Use local requirejs
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name sphinx_local_requirejs.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Avoid NaN crash in test_structural
Out-of-range frequency.cycle starting values make untransform_params return NaN,
which crashes summary()
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name fix_test_bounds.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Don't fail documentation build on examples exceptions
If example code outputs an error when run,
ipython (ipython:: blocks) now fails the whole build by default,
and tools/nbgenerate.py (.ipynb files) now leaves out that file
(leaving a broken link in the examples index).
As some examples use downloaded data and are hence expected to be
unrunnable on a Debian buildd, this instead keeps them
(::ipython with error output, .ipynb without output).
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name sphinx_ignore_errors.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Don't try to use a theme we don't have
Upstream switched to sphinx_material in
1d7aafad5c3ad634c9ae2122881c1809c2ccbcbe
and Debian doesn't have that theme.
(I haven't investigated packaging it, which might be a better long-term option.)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed (except maybe the version fix? or is that no longer needed at all?)
Gbp-Pq: Name no_sphinx_material.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Make the module reference tables appear
Our numpydoc is old enough to still need this manually set
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug: https://github.com/numpy/numpydoc/issues/141
Forwarded: not-needed
Gbp-Pq: Name sphinx_autosummary.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Xfail or loosen tolerances on flaky tests
The test_nondiagonal_obs_cov one hasn't yet been seen in Debian
but is known in upstream CI:
https://github.com/statsmodels/statsmodels/pull/6475#issuecomment-
581151072
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>, Graham Inggs
Bug-Debian: https://bugs.debian.org/944054
Forwarded: no
Gbp-Pq: Name 944054_flaky_tests.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Don't try to write to the source directory
Not allowed in autopkgtest
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name use_tmp_path.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Use Python 3 in script shebangs
(Skipping the two .pyx files with #!python - they need to be compiled
with Cython, not directly run with any version of plain Python)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name python3_shebangs.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Loosen tolerances on some tests on i386
Failure logs:
test_smoothing
https://buildd.debian.org/status/fetch.php?pkg=statsmodels&arch=i386&ver=0.9.0-3&stamp=
1567157609&raw=0
test_multivariate_switch_univariate
https://buildd.debian.org/status/fetch.php?pkg=statsmodels&arch=i386&ver=0.12.0-1&stamp=
1599693472&raw=0
As it only fails on i386 and isn't far wrong,
I suspect different rounding due to x87 excess precision
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/938949
Forwarded: no
Gbp-Pq: Name i386_loosen_test_tolerances.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Use example data from an R package we have
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use_available_data.patch
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
KalmanFilter broken on armhf / hppa
Warn on use, and xfail the tests
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/924036
Bug-Ubuntu: https://launchpad.net/bugs/
1819227
Forwarded: no
Gbp-Pq: Name xfail_kalman_armhf.patch
Diane Trout [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Use intersphinx inventory files from other Debian packages.
Forwarded: not-needed
Gbp-Pq: Name use-system-inventory
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
to prevent failed test on i386
Author: Yaroslav Halchenko <debian@onerussian.com>
Bug: https://github.com/statsmodels/statsmodels/issues/1831
Forwarded: upstream consider this probably fixed
Gbp-Pq: Name up_reduce_test_precision
Debian Science Maintainers [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Use cached datasets in tests and examples
Also remove a download that isn't actually used in that example.
This allows the tests to pass, and at least some of the examples to
be built, in an offline environment such as a Debian buildd.
The cached data is extracted from R packages by debian/datasets/*.
Author: Diane Trout <diane@ghic.org>, Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use-cached-datasets
Rebecca N. Palmer [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
statsmodels (0.12.0-2) unstable; urgency=medium
* Loosen some test tolerances on i386.
* Require scipy >= 1.5 to avoid autopkgtest failure.
[dgit import unpatched statsmodels 0.12.0-2]
Rebecca N. Palmer [Fri, 11 Sep 2020 06:24:40 +0000 (07:24 +0100)]
Import statsmodels_0.12.0-2.debian.tar.xz
[dgit import tarball statsmodels 0.12.0-2 statsmodels_0.12.0-2.debian.tar.xz]
Rebecca N. Palmer [Tue, 8 Sep 2020 17:35:08 +0000 (18:35 +0100)]
Import statsmodels_0.12.0.orig.tar.gz
[dgit import orig statsmodels_0.12.0.orig.tar.gz]
Raspbian automatic forward porter [Sat, 18 Jul 2020 01:48:48 +0000 (02:48 +0100)]
Merge version 0.11.1-2+rpi1 and 0.11.1-3 to produce 0.11.1-3+rpi1
Rebecca N. Palmer [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Merge statsmodels (0.11.1-3) import into refs/heads/workingbranch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Ignore new deprecation warnings
Origin: upstream
8873ab357c9a2a7d3eeeeb1cf1ac63d72a687a75
Bug-Debian: https://bugs.debian.org/964700
Author: Kevin Sheppard
Forwarded: not-needed
Gbp-Pq: Name numpy119_compat.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Remove code from Stack Overflow and wasteful loop
Stack Overflow content is CC-BY-SA licensed,
which this package is not supposed to be. This snippet may be
too small to be copyrightable, but removing it to be safe.
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name remove_ccbysa_snippet.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Allow tests to fail if multiprocessing is not available
This is currently the case on hurd-i386
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_no_multiprocessing.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
DynamicFactor test fail on ppc64el
Warn on use and xfail test
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_dynamicfactor_ppc64el.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Markov* broken on armhf
Warn on use, and xfail tests
I suspect https://github.com/statsmodels/statsmodels/pull/5826
but that improved it enough on amd64 that
reverting it isn't a good solution.
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_regimeswitching_armhf.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Don't require warnings that not all arches produce
numpy 0/0 (in tvalues): armel doesn't
scipy.stats NaN: Upstream already note i386 doesn't,
we find ppc64el and s390x don't either
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/956882
Forwarded: no
Gbp-Pq: Name dont_require_warnings.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Fix typos
Origin: lintian
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name spelling.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Use local requirejs
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name sphinx_local_requirejs.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Avoid NaN crash in test_structural
Out-of-range frequency.cycle starting values make untransform_params return NaN,
which crashes summary()
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name fix_test_bounds.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Don't fail documentation build on examples exceptions
If example code outputs an error when run,
ipython (ipython:: blocks) now fails the whole build by default,
and tools/nbgenerate.py (.ipynb files) now leaves out that file
(leaving a broken link in the examples index).
As some examples use downloaded data and are hence expected to be
unrunnable on a Debian buildd, this instead keeps them
(::ipython with error output, .ipynb without output).
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name sphinx_ignore_errors.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Don't try to use a theme we don't have
Upstream switched to sphinx_material in
1d7aafad5c3ad634c9ae2122881c1809c2ccbcbe
and Debian doesn't have that theme.
(I haven't investigated packaging it, which might be a better long-term option.)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed (except maybe the version fix? or is that no longer needed at all?)
Gbp-Pq: Name no_sphinx_material.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Include offset and exposure in GEEResults.fittedvalues
and remove misleading documentation from GLMResults.fittedvalues
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug: https://github.com/statsmodels/statsmodels/issues/5253
Forwarded: https://github.com/statsmodels/statsmodels/issues/5253#issuecomment-
578980169
Gbp-Pq: Name up5253_gee_offset.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Make the module reference tables appear
Our numpydoc is old enough to still need this manually set
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug: https://github.com/numpy/numpydoc/issues/141
Forwarded: not-needed
Gbp-Pq: Name sphinx_autosummary.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Xfail or loosen tolerances on flaky tests
The test_nondiagonal_obs_cov one hasn't yet been seen in Debian
but is known in upstream CI:
https://github.com/statsmodels/statsmodels/pull/6475#issuecomment-
581151072
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>, Graham Inggs
Bug-Debian: https://bugs.debian.org/944054
Forwarded: no
Gbp-Pq: Name 944054_flaky_tests.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Don't try to write to the source directory
Not allowed in autopkgtest
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name use_tmp_path.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Use Python 3 in script shebangs
(Skipping the two .pyx files with #!python - they need to be compiled
with Cython, not directly run with any version of plain Python)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name python3_shebangs.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Loosen tolerances on some tests on i386
Failure log
https://buildd.debian.org/status/fetch.php?pkg=statsmodels&arch=i386&ver=0.9.0-3&stamp=
1567157609&raw=0
As it only fails on i386 (not x32) and isn't far over tolerance,
I suspect different rounding due to x87 excess precision
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/938949
Forwarded: no
Gbp-Pq: Name i386_loosen_test_tolerances.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Use example data from an R package we have
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use_available_data.patch
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
KalmanFilter broken on armhf / hppa
Warn on use, and xfail the tests
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/924036
Bug-Ubuntu: https://launchpad.net/bugs/
1819227
Forwarded: no
Gbp-Pq: Name xfail_kalman_armhf.patch
Diane Trout [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Use intersphinx inventory files from other Debian packages.
Forwarded: not-needed
Gbp-Pq: Name use-system-inventory
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
to prevent failed test on i386
Author: Yaroslav Halchenko <debian@onerussian.com>
Bug: https://github.com/statsmodels/statsmodels/issues/1831
Forwarded: upstream consider this probably fixed
Gbp-Pq: Name up_reduce_test_precision
Debian Science Maintainers [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Use cached datasets in tests and examples
Also remove a download that isn't actually used in that example.
This allows the tests to pass, and at least some of the examples to
be built, in an offline environment such as a Debian buildd.
The cached data is extracted from R packages by debian/datasets/*.
Author: Diane Trout <diane@ghic.org>, Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use-cached-datasets
Rebecca N. Palmer [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
statsmodels (0.11.1-3) unstable; urgency=medium
* Tests: ignore harmless new warnings. Closes: #964700.
* Add matplotlib etc to autopkgtest Depends to skip fewer tests.
[dgit import unpatched statsmodels 0.11.1-3]
Rebecca N. Palmer [Fri, 10 Jul 2020 05:46:36 +0000 (06:46 +0100)]
Import statsmodels_0.11.1-3.debian.tar.xz
[dgit import tarball statsmodels 0.11.1-3 statsmodels_0.11.1-3.debian.tar.xz]
Peter Michael Green [Thu, 14 May 2020 19:20:30 +0000 (19:20 +0000)]
Disable documentation build due to hang.
Raspbian automatic forward porter [Wed, 13 May 2020 06:48:22 +0000 (07:48 +0100)]
Merge version 0.10.2-2+rpi1 and 0.11.1-2 to produce 0.11.1-2+rpi1
Rebecca N. Palmer [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Merge statsmodels (0.11.1-2) import into refs/heads/workingbranch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Remove code from Stack Overflow and wasteful loop
Stack Overflow content is CC-BY-SA licensed,
which this package is not supposed to be. This snippet may be
too small to be copyrightable, but removing it to be safe.
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name remove_ccbysa_snippet.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Allow tests to fail if multiprocessing is not available
This is currently the case on hurd-i386
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_no_multiprocessing.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
DynamicFactor test fail on ppc64el
Warn on use and xfail test
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_dynamicfactor_ppc64el.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Markov* broken on armhf
Warn on use, and xfail tests
I suspect https://github.com/statsmodels/statsmodels/pull/5826
but that improved it enough on amd64 that
reverting it isn't a good solution.
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name xfail_regimeswitching_armhf.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Don't require warnings that not all arches produce
numpy 0/0 (in tvalues): armel doesn't
scipy.stats NaN: Upstream already note i386 doesn't,
we find ppc64el and s390x don't either
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/956882
Forwarded: no
Gbp-Pq: Name dont_require_warnings.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Fix typos
Origin: lintian
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name spelling.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Use local requirejs
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name sphinx_local_requirejs.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Avoid NaN crash in test_structural
Out-of-range frequency.cycle starting values make untransform_params return NaN,
which crashes summary()
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name fix_test_bounds.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Don't fail documentation build on examples exceptions
If example code outputs an error when run,
ipython (ipython:: blocks) now fails the whole build by default,
and tools/nbgenerate.py (.ipynb files) now leaves out that file
(leaving a broken link in the examples index).
As some examples use downloaded data and are hence expected to be
unrunnable on a Debian buildd, this instead keeps them
(::ipython with error output, .ipynb without output).
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name sphinx_ignore_errors.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Don't try to use a theme we don't have
Upstream switched to sphinx_material in
1d7aafad5c3ad634c9ae2122881c1809c2ccbcbe
and Debian doesn't have that theme.
(I haven't investigated packaging it, which might be a better long-term option.)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed (except maybe the version fix? or is that no longer needed at all?)
Gbp-Pq: Name no_sphinx_material.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Include offset and exposure in GEEResults.fittedvalues
and remove misleading documentation from GLMResults.fittedvalues
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug: https://github.com/statsmodels/statsmodels/issues/5253
Forwarded: https://github.com/statsmodels/statsmodels/issues/5253#issuecomment-
578980169
Gbp-Pq: Name up5253_gee_offset.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Make the module reference tables appear
Our numpydoc is old enough to still need this manually set
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug: https://github.com/numpy/numpydoc/issues/141
Forwarded: not-needed
Gbp-Pq: Name sphinx_autosummary.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Xfail or loosen tolerances on flaky tests
The test_nondiagonal_obs_cov one hasn't yet been seen in Debian
but is known in upstream CI:
https://github.com/statsmodels/statsmodels/pull/6475#issuecomment-
581151072
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>, Graham Inggs
Bug-Debian: https://bugs.debian.org/944054
Forwarded: no
Gbp-Pq: Name 944054_flaky_tests.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Don't try to write to the source directory
Not allowed in autopkgtest
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name use_tmp_path.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Use Python 3 in script shebangs
(Skipping the two .pyx files with #!python - they need to be compiled
with Cython, not directly run with any version of plain Python)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name python3_shebangs.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Loosen tolerances on some tests on i386
Failure log
https://buildd.debian.org/status/fetch.php?pkg=statsmodels&arch=i386&ver=0.9.0-3&stamp=
1567157609&raw=0
As it only fails on i386 (not x32) and isn't far over tolerance,
I suspect different rounding due to x87 excess precision
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/938949
Forwarded: no
Gbp-Pq: Name i386_loosen_test_tolerances.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Use example data from an R package we have
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use_available_data.patch
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
KalmanFilter broken on armhf / hppa
Warn on use, and xfail the tests
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/924036
Bug-Ubuntu: https://launchpad.net/bugs/
1819227
Forwarded: no
Gbp-Pq: Name xfail_kalman_armhf.patch
Diane Trout [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Use intersphinx inventory files from other Debian packages.
Forwarded: not-needed
Gbp-Pq: Name use-system-inventory
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
to prevent failed test on i386
Author: Yaroslav Halchenko <debian@onerussian.com>
Bug: https://github.com/statsmodels/statsmodels/issues/1831
Forwarded: upstream consider this probably fixed
Gbp-Pq: Name up_reduce_test_precision
Debian Science Maintainers [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Use cached datasets in tests and examples
Also remove a download that isn't actually used in that example.
This allows the tests to pass, and at least some of the examples to
be built, in an offline environment such as a Debian buildd.
The cached data is extracted from R packages by debian/datasets/*.
Author: Diane Trout <diane@ghic.org>, Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use-cached-datasets
Rebecca N. Palmer [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
statsmodels (0.11.1-2) unstable; urgency=medium
* Tests: don't require a numpy warning that doesn't exist on armel.
Closes: #956882.
* Tests: allow no-multiprocessing failures (hurd-i386).
* Remove inconveniently licensed (CC-BY-SA) snippet.
[dgit import unpatched statsmodels 0.11.1-2]
Rebecca N. Palmer [Fri, 24 Apr 2020 14:17:04 +0000 (15:17 +0100)]
Import statsmodels_0.11.1-2.debian.tar.xz
[dgit import tarball statsmodels 0.11.1-2 statsmodels_0.11.1-2.debian.tar.xz]
Rebecca N. Palmer [Mon, 24 Feb 2020 22:49:06 +0000 (22:49 +0000)]
Import statsmodels_0.11.1.orig.tar.gz
[dgit import orig statsmodels_0.11.1.orig.tar.gz]
Peter Michael Green [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Merge statsmodels (0.10.2-2+rpi1) import into refs/heads/workingbranch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Include offset and exposure in GEEResults.fittedvalues
and remove misleading documentation from GLMResults.fittedvalues
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug: https://github.com/statsmodels/statsmodels/issues/5253
Forwarded: https://github.com/statsmodels/statsmodels/issues/5253#issuecomment-
578980169
Gbp-Pq: Name up5253_gee_offset.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Don't fail documentation build on examples exceptions
ipython now fails the whole build by default if anything outputs an error,
which some of ours do when they can't download data.
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name sphinx_ignore_errors.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Make the module reference tables appear
Our numpydoc is old enough to still need this manually set
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug: https://github.com/numpy/numpydoc/issues/141
Forwarded: not-needed
Gbp-Pq: Name sphinx_autosummary.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Xfail or loosen tolerances on flaky tests
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>, Graham Inggs
Bug-Debian: https://bugs.debian.org/944054
Forwarded: partially https://github.com/statsmodels/statsmodels/pull/6222
Gbp-Pq: Name 944054_flaky_tests.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Use same Python version for subprocess test
Avoids a fail due to mismatched extension .so version
when multiple Python 3 versions are available
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name subprocess_use_same_pyver.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Don't try to write to the source directory
Not allowed in autopkgtest
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name use_tmp_path.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Use Python 3 in script shebangs
(Skipping the two .pyx files with #!python - they need to be compiled
with Cython, not directly run with any version of plain Python)
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name python3_shebangs.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Use scipy.special not scipy.misc
factorial, comb and logsumexp have been moved: trying to import them
from the old location is an error in scipy 1.3+.
https://ci.debian.net/data/autopkgtest/unstable/amd64/s/statsmodels/
2949940/log.gz
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: accepted https://github.com/statsmodels/statsmodels/pull/6164
Gbp-Pq: Name scipy13_compat.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Loosen tolerances on some tests on i386
Failure log
https://buildd.debian.org/status/fetch.php?pkg=statsmodels&arch=i386&ver=0.9.0-3&stamp=
1567157609&raw=0
As it only fails on i386 (not x32) and isn't far over tolerance,
I suspect different rounding due to x87 excess precision
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/938949
Forwarded: no
Gbp-Pq: Name i386_loosen_test_tolerances.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Warn that load_pickle() etc are for trusted data only
https://docs.python.org/3/library/pickle.html#restricting-globals
Author: Rebecca N. Palmer
Forwarded: https://github.com/statsmodels/statsmodels/pull/6162
Gbp-Pq: Name cache_security.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Use example data from an R package we have
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use_available_data.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
KalmanFilter broken on armhf / hppa
Warn on use, and xfail the tests
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Bug-Debian: https://bugs.debian.org/924036
Bug-Ubuntu: https://launchpad.net/bugs/
1819227
Forwarded: no
Gbp-Pq: Name xfail_kalman_armhf.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Give a more useful error if cvxopt is not found
message wasn't actually passed to anything, so it previously failed at
statsmodels/base/optimizer.py:13 (_check_method) with
"Unknown fit method l1_cvxopt_cp"
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: accepted https://github.com/statsmodels/statsmodels/pull/6163
Gbp-Pq: Name show_cvxopt_error.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Change some more links to https
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: https://github.com/statsmodels/statsmodels/pull/5937
Gbp-Pq: Name link_security2.patch
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Don't fetch Twitter logo
Author: Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: no
Gbp-Pq: Name privacy.patch
Diane Trout [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Use intersphinx inventory files from other Debian packages.
Forwarded: not-needed
Gbp-Pq: Name use-system-inventory
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
to prevent failed test on i386
Author: Yaroslav Halchenko <debian@onerussian.com>
Bug: https://github.com/statsmodels/statsmodels/issues/1831
Forwarded: upstream consider this probably fixed
Gbp-Pq: Name up_reduce_test_precision
Debian Science Maintainers [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Use cached datasets for building documentation/examples
Also remove a download that isn't actually used in that example.
This allows at least some of the examples to be built in an
offline environment such as a Debian buildd.
The cached data is extracted from R packages by debian/datasets/*.
Author: Diane Trout <diane@ghic.org>, Rebecca N. Palmer <rebecca_palmer@zoho.com>
Forwarded: not-needed
Gbp-Pq: Name use-cached-datasets
Peter Michael Green [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
statsmodels (0.10.2-2+rpi1) bullseye-staging; urgency=medium
* Disable testsuite.
[dgit import unpatched statsmodels 0.10.2-2+rpi1]
Peter Michael Green [Tue, 4 Feb 2020 00:34:03 +0000 (00:34 +0000)]
Import statsmodels_0.10.2-2+rpi1.debian.tar.xz
[dgit import tarball statsmodels 0.10.2-2+rpi1 statsmodels_0.10.2-2+rpi1.debian.tar.xz]
Rebecca N. Palmer [Tue, 28 Jan 2020 22:29:29 +0000 (22:29 +0000)]
Merge statsmodels (0.10.2-2) import into refs/heads/workingbranch