Bug#920711: r-cran-repr: autopkgtest regression

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Bug#920711: r-cran-repr: autopkgtest regression

Graham Inggs-3
Source: r-cran-repr
Version: 0.19.1-1
Severity: serious
X-Debbugs-CC: [hidden email]
User: [hidden email]
Usertags: regression

Hi Maintainer

Since the upload of r-cran-tibble 2.0.0-1 to unstable, r-cran-repr has
been failing its autopkgtests [1] with the following error:

 > test_check('repr')
-- 1. Failure: ellip_limit_arr preserves rownames when limiting rows
(@test_arra
`limited_tbl` not identical to `expected_df_mat`.
Attributes: < Component "dimnames": Component 1: 4 string mismatches >

-- 2. Failure: ellip_limit_arr preserves rownames when limiting rows
(@test_arra
`limited_tbl` not identical to `expected_df_mat`.
Attributes: < Component "dimnames": Component 1: 5 string mismatches >

== testthat results
===========================================================
OK: 128 SKIPPED: 5 FAILED: 2
1. Failure: ellip_limit_arr preserves rownames when limiting rows
(@test_array_manipulation.r#229)
2. Failure: ellip_limit_arr preserves rownames when limiting rows
(@test_array_manipulation.r#251)

Error: testthat unit tests failed

Regards
Graham


[1] https://ci.debian.net/packages/r/r-cran-repr/unstable/amd64/

Reply | Threaded
Open this post in threaded view
|

Bug#920711: Test suite seems to uncover conflict between new version of tibble and repr (Was: Bug#920711: r-cran-repr: autopkgtest regression)

Andreas Tille-2
Hi Philipp,

the continuous integration test in Debian has uncovered an issue after
tibble 2.0.0 was uploaded.  Please have a look below.  The full build
log can be found here:

   https://ci.debian.net/data/autopkgtest/unstable/amd64/r/r-cran-repr/1847462/log.gz

(Please scroll down to the very bottom of that page!)

Any hint how this can be solved?

Kind regards

       Andreas.

On Mon, Jan 28, 2019 at 04:24:10PM +0200, Graham Inggs wrote:

> Source: r-cran-repr
> Version: 0.19.1-1
> Severity: serious
> X-Debbugs-CC: [hidden email]
> User: [hidden email]
> Usertags: regression
>
> Hi Maintainer
>
> Since the upload of r-cran-tibble 2.0.0-1 to unstable, r-cran-repr has been
> failing its autopkgtests [1] with the following error:
>
> > test_check('repr')
> -- 1. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_arra
> `limited_tbl` not identical to `expected_df_mat`.
> Attributes: < Component "dimnames": Component 1: 4 string mismatches >
>
> -- 2. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_arra
> `limited_tbl` not identical to `expected_df_mat`.
> Attributes: < Component "dimnames": Component 1: 5 string mismatches >
>
> == testthat results
> ===========================================================
> OK: 128 SKIPPED: 5 FAILED: 2
> 1. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_array_manipulation.r#229)
> 2. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_array_manipulation.r#251)
>
> Error: testthat unit tests failed
>
> Regards
> Graham
>
>
> [1] https://ci.debian.net/packages/r/r-cran-repr/unstable/amd64/
>
> _______________________________________________
> R-pkg-team mailing list
> [hidden email]
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/r-pkg-team

--
http://fam-tille.de

Reply | Threaded
Open this post in threaded view
|

Bug#920711: Test suite seems to uncover conflict between new version of tibble and repr (Was: Bug#920711: r-cran-repr: autopkgtest regression)

Philipp Angerer
Hi! Umm, why me? I’m neither the tibble author nor a Debian package maintainer :D

Andreas Tille <[hidden email]> schrieb am Di., 5. Feb. 2019 um 10:36 Uhr:
Hi Philipp,

the continuous integration test in Debian has uncovered an issue after
tibble 2.0.0 was uploaded.  Please have a look below.  The full build
log can be found here:

   https://ci.debian.net/data/autopkgtest/unstable/amd64/r/r-cran-repr/1847462/log.gz

(Please scroll down to the very bottom of that page!)

Any hint how this can be solved?

Kind regards

       Andreas.

On Mon, Jan 28, 2019 at 04:24:10PM +0200, Graham Inggs wrote:
> Source: r-cran-repr
> Version: 0.19.1-1
> Severity: serious
> X-Debbugs-CC: [hidden email]
> User: [hidden email]
> Usertags: regression
>
> Hi Maintainer
>
> Since the upload of r-cran-tibble 2.0.0-1 to unstable, r-cran-repr has been
> failing its autopkgtests [1] with the following error:
>
> > test_check('repr')
> -- 1. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_arra
> `limited_tbl` not identical to `expected_df_mat`.
> Attributes: < Component "dimnames": Component 1: 4 string mismatches >
>
> -- 2. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_arra
> `limited_tbl` not identical to `expected_df_mat`.
> Attributes: < Component "dimnames": Component 1: 5 string mismatches >
>
> == testthat results
> ===========================================================
> OK: 128 SKIPPED: 5 FAILED: 2
> 1. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_array_manipulation.r#229)
> 2. Failure: ellip_limit_arr preserves rownames when limiting rows
> (@test_array_manipulation.r#251)
>
> Error: testthat unit tests failed
>
> Regards
> Graham
>
>
> [1] https://ci.debian.net/packages/r/r-cran-repr/unstable/amd64/
>
> _______________________________________________
> R-pkg-team mailing list
> [hidden email]
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/r-pkg-team

--
http://fam-tille.de
Reply | Threaded
Open this post in threaded view
|

Bug#920711: Test suite seems to uncover conflict between new version of tibble and repr (Was: Bug#920711: r-cran-repr: autopkgtest regression)

Graham Inggs-3
In reply to this post by Graham Inggs-3
Hi!

On Tue, 5 Feb 2019 at 22:34, Andreas Tille <[hidden email]> wrote:
> I can also add r-cran-tibble via
>
>     apt-get install --no-install-recommends r-cran-tibble
>
> and the test suite keeps on succeeding!  However, after intalling
> r-cran-dplyr the described error occures while after deinstalling
> r-cran-dplyr the test suite passes again.  So the incompatibility
> seems to be rather between r-cran-repr and r-cran-dplyr.

The reason the two tests only fail when r-cran-dplyr is installed is because of:
    if (has_dplyr) { }
in both cases [1][2].

Regards
Graham


[1] https://salsa.debian.org/r-pkg-team/r-cran-repr/blob/master/tests/testthat/test_array_manipulation.r#L226
[2] https://salsa.debian.org/r-pkg-team/r-cran-repr/blob/master/tests/testthat/test_array_manipulation.r#L249

Reply | Threaded
Open this post in threaded view
|

Bug#920711: Test suite seems to uncover conflict between new version of tibble and repr (Was: Bug#920711: r-cran-repr: autopkgtest regression)

Andreas Tille-2
In reply to this post by Graham Inggs-3
Cool!  Thanks a lot, Andreas.

On Wed, Feb 06, 2019 at 01:05:51PM +0100, Philipp Angerer wrote:

> Oh whoops, this is about repr, not tibble! Should have read more carefully,
> sorry!
>
> I just tagged 0.19.2 and uploaded it to CRAN (should be up there soon),
> which works around the tibble-induced breakage:
> https://github.com/IRkernel/repr/releases/tag/0.19.2
>
> FYI: tibble has a schizophrenic approach to data.frames: It claims to be a
> subclass, but doesn’t properly support rownames (by design). Of course that
> has its reasons, but for code expecting rownames to be handles, this breaks
> things. And as they proceedingly remove rowname support from all
> operations, proceedingly more stuff breaks.
>
> Best, Philipp
>
> Andreas Tille <[hidden email]> schrieb am Mi., 6. Feb. 2019 um 11:28 Uhr:
>
> > Control: tags -1 help
> >
> > On Tue, Feb 05, 2019 at 10:56:11PM +0200, Graham Inggs wrote:
> > > Hi!
> > >
> > > On Tue, 5 Feb 2019 at 22:34, Andreas Tille <[hidden email]> wrote:
> > > > I can also add r-cran-tibble via
> > > >
> > > >     apt-get install --no-install-recommends r-cran-tibble
> > > >
> > > > and the test suite keeps on succeeding!  However, after intalling
> > > > r-cran-dplyr the described error occures while after deinstalling
> > > > r-cran-dplyr the test suite passes again.  So the incompatibility
> > > > seems to be rather between r-cran-repr and r-cran-dplyr.
> > >
> > > The reason the two tests only fail when r-cran-dplyr is installed is
> > because of:
> > >     if (has_dplyr) { }
> > > in both cases [1][2].
> >
> > But this does not explain why r-cran-repr fails after a new
> > r-cran-tibble is uploaded.  It remains a bit mysterious to me.  So I'm
> > tagging this bug "help" but I hesitate for tags upstream or a forwarded
> > field since I'm not sure what upstream to contact about this.
> >
> > Kind regards
> >
> >       Andreas.
> >
> > --
> > http://fam-tille.de
> >

> _______________________________________________
> R-pkg-team mailing list
> [hidden email]
> https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/r-pkg-team


--
http://fam-tille.de