Bug#770730: pre-approval for t-p-u upload/unblock: statsmodels/0.4.2-1.1
retitle 770730 unblock: statsmodels/0.4.2-1.2 tags 770730 -moreinfo thanks On Wed, Nov 26, 2014 at 11:59:08PM +0100, Michael Banck wrote: On Wed, Nov 26, 2014 at 10:38:08PM +, Jonathan Wiltshire wrote: I see it's been uploaded, but FTBFS on i386. Yeah - I tried it on an s390x porterbox (as later upstream releases FTBFS on that and powerpc), but didn't fathom it might break on i386. Tomasz has now updated his patch for i386 and I will sponsor an upload for that tomorrow. I have now uploaded -1.2 to testing, fixing the FTBFS on i386. Debdiff to the current version in testing is attached for unblock review. For the debdiff between -1.1 and -1.2, see https://bugs.debian.org/cgi-bin/bugreport.cgi?msg=100;filename=debdiff;att=1;bug=768695 Michael diff -Nru statsmodels-0.4.2/debian/changelog statsmodels-0.4.2/debian/changelog --- statsmodels-0.4.2/debian/changelog 2012-06-29 23:26:49.0 +0200 +++ statsmodels-0.4.2/debian/changelog 2014-11-29 16:33:12.0 +0100 @@ -1,3 +1,22 @@ +statsmodels (0.4.2-1.2) testing; urgency=medium + + * Non-maintainer upload, incorporating changes by Tomasz Buchert. + * debian/patches/scipy-rvs-interface.patch: Removed again and ... + * debian/patches/testsuite-fixes.patch: ... folded into this patch, just +raising a nose.SkipTest() for the problematic test case. + * debian/patches/testsuite-fixes.patch: Amended for test_discrete.py, also +skipping this test case via raising a nose.SkipTest(), as it leads to a +FTBFS error on i386 (Closes: #768695). + + -- Michael Banck mba...@debian.org Sat, 29 Nov 2014 16:19:51 +0100 + +statsmodels (0.4.2-1.1) testing; urgency=medium + + * Non-maintainer upload. + * Fixes various problems with build process (Closes: #768695) + + -- Tomasz Buchert tomasz.buch...@inria.fr Sun, 23 Nov 2014 17:46:48 +0100 + statsmodels (0.4.2-1) unstable; urgency=low * Fresh upstream release addressing FTBFS across big-endian architectures. diff -Nru statsmodels-0.4.2/debian/patches/series statsmodels-0.4.2/debian/patches/series --- statsmodels-0.4.2/debian/patches/series 2012-06-29 23:26:49.0 +0200 +++ statsmodels-0.4.2/debian/patches/series 2014-11-29 16:40:05.0 +0100 @@ -0,0 +1,2 @@ +sphinx-ipython.patch +testsuite-fixes.patch diff -Nru statsmodels-0.4.2/debian/patches/sphinx-ipython.patch statsmodels-0.4.2/debian/patches/sphinx-ipython.patch --- statsmodels-0.4.2/debian/patches/sphinx-ipython.patch 1970-01-01 01:00:00.0 +0100 +++ statsmodels-0.4.2/debian/patches/sphinx-ipython.patch 2014-11-29 14:08:36.0 +0100 @@ -0,0 +1,14 @@ +Description: Fix building of docs + See https://github.com/matplotlib/matplotlib/issues/2967 for more info. + +--- a/docs/source/conf.py b/docs/source/conf.py +@@ -33,7 +33,7 @@ + 'matplotlib.sphinxext.plot_directive', + 'matplotlib.sphinxext.only_directives', + 'ipython_console_highlighting', +- 'ipython_directive', ++ 'IPython.sphinxext.ipython_directive', + 'numpy_ext.numpydoc'] + + # plot_directive is broken on old matplotlib diff -Nru statsmodels-0.4.2/debian/patches/testsuite-fixes.patch statsmodels-0.4.2/debian/patches/testsuite-fixes.patch --- statsmodels-0.4.2/debian/patches/testsuite-fixes.patch 1970-01-01 01:00:00.0 +0100 +++ statsmodels-0.4.2/debian/patches/testsuite-fixes.patch 2014-11-29 14:08:36.0 +0100 @@ -0,0 +1,165 @@ +Description: Fix various testsuite problems + The testsuite depends on version-specific functionality + of various dependencies like numpy or scipy. This patch fixes + problems caused by versions in jessie release of these dependencies. + . + statsmodels/tools/tools.py: + = unexisting attribute in numpy object + statsmodels/sandbox/distributions/tests/testtransf.py: + statsmodels/tsa/filters/tests/test_filters.py: + = scipy interface incompatibilities + statsmodels/tsa/base/tests/test_datetools.py: + = DateRange class is not present in jessie pandas + statsmodels/sandbox/distributions/extras.py: + = a mistake fixed in newer releases of statsmodels + statsmodels/sandbox/tests/test_gam.py: + = incompatible scipy interface for rvs method + statsmodels/discrete/tests/test_discrete.py: + = failures due to differences between architectures, see + https://github.com/statsmodels/statsmodels/commit/ca701e7a + +--- a/statsmodels/tools/tools.py b/statsmodels/tools/tools.py +@@ -231,7 +231,7 @@ + + def _series_add_constant(data, prepend): + const = np.ones_like(data) +-const.name = 'const' ++# const.name = 'const' + if not prepend: + results = DataFrame([data, const]).T + results.columns = [data.name, 'const'] +--- a/statsmodels/sandbox/distributions/tests/testtransf.py b/statsmodels/sandbox/distributions/tests/testtransf.py +@@ -88,8 +88,8 @@ + (absnormalg, stats.halfnorm), + (absnormalg,
Bug#770730: pre-approval for t-p-u upload/unblock: statsmodels/0.4.2-1.1
Control: tag -1 moreinfo On Sun, Nov 23, 2014 at 05:42:19PM +, Adam D. Barratt wrote: Control: tags -1 + confirmed moreinfo On Sun, 2014-11-23 at 18:23 +0100, Michael Banck wrote: Please pre-approve an upload statsmodels to testing-proposed-updates in order to fix #768695. As unstable has a new upstream version, this needs to be fixed in jessie. The NMU debdiff I am going to sponsor is attached. Thanks for this. Please go ahead, and remove the moreinfo tag once the upload has been accepted. One comment: + interface. We remove these failing tests altogother. s/altogother/altogether/ I see it's been uploaded, but FTBFS on i386. -- Jonathan Wiltshire j...@debian.org Debian Developer http://people.debian.org/~jmw 4096R: 0xD3524C51 / 0A55 B7C5 1223 3942 86EC 74C3 5394 479D D352 4C51 signature.asc Description: Digital signature
Bug#770730: pre-approval for t-p-u upload/unblock: statsmodels/0.4.2-1.1
On Wed, Nov 26, 2014 at 10:38:08PM +, Jonathan Wiltshire wrote: I see it's been uploaded, but FTBFS on i386. Yeah - I tried it on an s390x porterbox (as later upstream releases FTBFS on that and powerpc), but didn't fathom it might break on i386. Tomasz has now updated his patch for i386 and I will sponsor an upload for that tomorrow. Michael -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#770730: pre-approval for t-p-u upload/unblock: statsmodels/0.4.2-1.1
Package: release.debian.org Severity: normal User: release.debian@packages.debian.org Usertags: unblock Please pre-approve an upload statsmodels to testing-proposed-updates in order to fix #768695. As unstable has a new upstream version, this needs to be fixed in jessie. The NMU debdiff I am going to sponsor is attached. unblock statsmodels/0.4.2-1.1 -- System Information: Debian Release: jessie/sid APT prefers testing-updates APT policy: (500, 'testing-updates'), (500, 'testing') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 3.16-2-amd64 (SMP w/4 CPU cores) Locale: LANG=en_US.utf8, LC_CTYPE=en_US.utf8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash diff -Nru statsmodels-0.4.2/debian/changelog statsmodels-0.4.2/debian/changelog --- statsmodels-0.4.2/debian/changelog 2012-06-29 23:26:49.0 +0200 +++ statsmodels-0.4.2/debian/changelog 2014-11-23 17:55:26.0 +0100 @@ -1,3 +1,10 @@ +statsmodels (0.4.2-1.1) testing; urgency=medium + + * Non-maintainer upload. + * Fixes various problems with build process + + -- Tomasz Buchert tomasz.buch...@inria.fr Sun, 23 Nov 2014 17:46:48 +0100 + statsmodels (0.4.2-1) unstable; urgency=low * Fresh upstream release addressing FTBFS across big-endian architectures. diff -Nru statsmodels-0.4.2/debian/patches/scipy-rvs-interface.patch statsmodels-0.4.2/debian/patches/scipy-rvs-interface.patch --- statsmodels-0.4.2/debian/patches/scipy-rvs-interface.patch 1970-01-01 01:00:00.0 +0100 +++ statsmodels-0.4.2/debian/patches/scipy-rvs-interface.patch 2014-11-23 17:57:08.0 +0100 @@ -0,0 +1,128 @@ +Description: remove tests that use uncompatible scipy interface + The tests use an rvs method which has an incompatible + interface. We remove these failing tests altogother. + +--- statsmodels-0.4.2.orig/statsmodels/sandbox/tests/test_gam.py statsmodels-0.4.2/statsmodels/sandbox/tests/test_gam.py +@@ -187,121 +187,3 @@ class TestAdditiveModel(BaseAM, CheckAM) + const = res_gam.alpha + sum([ss.params[1] for ss in m.smoothers]) + #print const, slopes + res1.params = np.array([const] + slopes) +- +- +-class BaseGAM(BaseAM, CheckGAM): +- +-def init(self): +-nobs = self.nobs +-y_true, x, exog = self.y_true, self.x, self.exog +-if not hasattr(self, 'scale'): +-scale = 1 +-else: +-scale = self.scale +- +-f = self.family +- +-self.mu_true = mu_true = f.link.inverse(y_true) +- +-np.random.seed(8765993) +-#y_obs = np.asarray([stats.poisson.rvs(p) for p in mu], float) +-y_obs = self.rvs(mu_true, scale=scale, size=nobs) #this should work +-m = GAM(y_obs, x, family=f) #TODO: y_obs is twice __init__ and fit +-m.fit(y_obs, maxiter=100) +-res_gam = m.results +-self.res_gam = res_gam #attached for debugging +-self.mod_gam = m #attached for debugging +- +-res_glm = GLM(y_obs, exog, family=f).fit() +- +-#Note: there still are some naming inconsistencies +-self.res1 = res1 = Dummy() #for gam model +-#res2 = Dummy() #for benchmark +-self.res2 = res2 = res_glm #reuse existing glm results, will add additional +- +-#eta in GLM terminology +-res2.y_pred = res_glm.model.predict(res_glm.params, exog, linear=True) +-res1.y_pred = res_gam.predict(x) +-res1.y_predshort = res_gam.predict(x[:10]) #, linear=True) +- +-#mu +-res2.mu_pred = res_glm.model.predict(res_glm.params, exog, linear=False) +-res1.mu_pred = res_gam.mu +- +-#parameters +-slopes = [i for ss in m.smoothers for i in ss.params[1:]] +-const = res_gam.alpha + sum([ss.params[1] for ss in m.smoothers]) +-res1.params = np.array([const] + slopes) +- +- +-class TestGAMPoisson(BaseGAM): +- +-def __init__(self): +-super(self.__class__, self).__init__() #initialize DGP +- +-self.family = family.Poisson() +-self.rvs = stats.poisson.rvs +- +-self.init() +- +-class TestGAMBinomial(BaseGAM): +- +-def __init__(self): +-super(self.__class__, self).__init__() #initialize DGP +- +-self.family = family.Binomial() +-self.rvs = stats.bernoulli.rvs +- +-self.init() +- +-class _estGAMGaussianLogLink(BaseGAM): +-#test failure, but maybe precision issue, not far off +-# np.mean(np.abs(tt.res2.mu_pred - tt.mu_true)) +-#0.80409736263199649 +-# np.mean(np.abs(tt.res2.mu_pred - tt.mu_true))/tt.mu_true.mean() +-#0.023258245077813208 +-# np.mean((tt.res2.mu_pred - tt.mu_true)**2)/tt.mu_true.mean() +-#0.022989403735692578 +- +-def __init__(self): +-super(self.__class__, self).__init__() #initialize DGP +- +-self.family = family.Gaussian(links.log) +-self.rvs = stats.norm.rvs +-self.scale = 5 +- +-self.init() +- +- +-class TestGAMGamma(BaseGAM): +- +-def
Bug#770730: pre-approval for t-p-u upload/unblock: statsmodels/0.4.2-1.1
Control: tags -1 + confirmed moreinfo On Sun, 2014-11-23 at 18:23 +0100, Michael Banck wrote: Please pre-approve an upload statsmodels to testing-proposed-updates in order to fix #768695. As unstable has a new upstream version, this needs to be fixed in jessie. The NMU debdiff I am going to sponsor is attached. Thanks for this. Please go ahead, and remove the moreinfo tag once the upload has been accepted. One comment: + interface. We remove these failing tests altogother. s/altogother/altogether/ Regards, Adam -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
