Release 2.5-4 before Feb 8, 2019

[jarioksa]

Kurt Hornik of CRAN sent email saying:

Dear maintainer,

Please see the problems shown on
https://cran.r-project.org/web/checks/check_results_vegan.html.

Please correct before 2019-02-08 to safely retain your package on CRAN.

The breakage is the change of formula in R (issue #299) that we have already patched in the cran-2.5 branch.

There are not so many changes in this release, but this would also fix #303, #304 and the goodness problem that @legendre reported in email. adonis NA handling (issue #302) cannot be merged to this minor release: master branch has changed too much, and it must wait for the 2.6 releases.

I have also flagged #295 as a milestone issue that should be solved for this release. This issue is about handling partial ordination models in RsquareAdj. This is technically easy: we just revert some of the commits that removed R_adj² for partial models. It is only about deciding which commits should be reverted: RDA is obvious, but how to handle CCA?

[jarioksa]

Running win-builder gives an error that I can diagnose but I cannot trigger in my system. It is about if() conditions being longer than length 1. This used to be a warning, but win-builder makes it an error. The problem is that I cannot turn this into an error in my system and I cannot detect these cases and correct them. I have set R environment variable _R_CHECK_LENGTH_1_CONDITION_='TRUE' (with TRUE alternatively lower case, with or without quotes). This sets the environment OK when I use vector conditions in command line (such as if(runif(10) > 0.5) TRUE): This gives an error like should with this setting. However, when I run example(cca) with this problem, everything goes smoothly, and so goes R CMD check.

This error is not reported in the CRAN check page https://cran.r-project.org/web/checks/check_results_vegan.html, but I think CRAN submission check will find this. Therefore this should be fixed before the release. The only way I can proceed is to fix the check error win-builder reports, make a new package, submit it to win-builder, see where it fails and loop again. So if @gavinsimpson @psolymos or anybody can trigger the error, please have a look a the fixes (which usually are very simple after you find them). If you try checking these issues, remember to git fetch the current vegan tree, where I have fixed two cases.

Here an excerpt from the win-builder test which comes from example(cca):

* checking examples ...
** running examples for arch 'i386' ... ERROR
Running examples in 'vegan-Ex.R' failed
The error most likely occurred in:

> base::assign(".ptime", proc.time(), pos = "CheckExEnv")
> ### Name: cca
> ### Title: [Partial] [Constrained] Correspondence Analysis and Redundancy
> ###   Analysis
> ### Aliases: cca cca.default cca.formula rda rda.default rda.formula
> ### Keywords: multivariate
> 
> ### ** Examples
> 
> data(varespec)
> data(varechem)
> ## Common but bad way: use all variables you happen to have in your
> ## environmental data matrix
> vare.cca <- cca(varespec, varechem)
> vare.cca
Call: cca(X = varespec, Y = varechem)

              Inertia Proportion Rank
Total          2.0832     1.0000     
Constrained    1.4415     0.6920   14
Unconstrained  0.6417     0.3080    9
Inertia is scaled Chi-square 

Eigenvalues for constrained axes:
  CCA1   CCA2   CCA3   CCA4   CCA5   CCA6   CCA7   CCA8   CCA9  CCA10  CCA11 
0.4389 0.2918 0.1628 0.1421 0.1180 0.0890 0.0703 0.0584 0.0311 0.0133 0.0084 
 CCA12  CCA13  CCA14 
0.0065 0.0062 0.0047 

Eigenvalues for unconstrained axes:
    CA1     CA2     CA3     CA4     CA5     CA6     CA7     CA8     CA9 
0.19776 0.14193 0.10117 0.07079 0.05330 0.03330 0.01887 0.01510 0.00949 

> plot(vare.cca)
> ## Formula interface and a better model
> vare.cca <- cca(varespec ~ Al + P*(K + Baresoil), data=varechem)
> vare.cca
Call: cca(formula = varespec ~ Al + P * (K + Baresoil), data =
varechem)

              Inertia Proportion Rank
Total           2.083      1.000     
Constrained     1.046      0.502    6
Unconstrained   1.038      0.498   17
Inertia is scaled Chi-square 

Eigenvalues for constrained axes:
  CCA1   CCA2   CCA3   CCA4   CCA5   CCA6 
0.3756 0.2342 0.1407 0.1323 0.1068 0.0561 

Eigenvalues for unconstrained axes:
    CA1     CA2     CA3     CA4     CA5     CA6     CA7     CA8 
0.27577 0.15411 0.13536 0.11803 0.08887 0.05511 0.04919 0.03781 
(Showing 8 of 17 unconstrained eigenvalues)

> plot(vare.cca)
> ## Partialling out and negative components of variance
> cca(varespec ~ Ca, varechem)
Call: cca(formula = varespec ~ Ca, data = varechem)

              Inertia Proportion Rank
Total         2.08320    1.00000     
Constrained   0.15722    0.07547    1
Unconstrained 1.92598    0.92453   22
Inertia is scaled Chi-square 

Eigenvalues for constrained axes:
   CCA1 
0.15722 

Eigenvalues for unconstrained axes:
   CA1    CA2    CA3    CA4    CA5    CA6    CA7    CA8 
0.4745 0.2939 0.2140 0.1954 0.1748 0.1171 0.1121 0.0880 
(Showing 8 of 22 unconstrained eigenvalues)

> cca(varespec ~ Ca + Condition(pH), varechem)
Call: cca(formula = varespec ~ Ca + Condition(pH), data = varechem)

              Inertia Proportion Rank
Total          2.0832     1.0000     
Conditional    0.1458     0.0700    1
Constrained    0.1827     0.0877    1
Unconstrained  1.7547     0.8423   21
Inertia is scaled Chi-square 

Eigenvalues for constrained axes:
   CCA1 
0.18269 

Eigenvalues for unconstrained axes:
   CA1    CA2    CA3    CA4    CA5    CA6    CA7    CA8 
0.3834 0.2749 0.2123 0.1760 0.1701 0.1161 0.1089 0.0880 
(Showing 8 of 21 unconstrained eigenvalues)

> ## RDA
> data(dune)
> data(dune.env)
> dune.Manure <- rda(dune ~ Manure, dune.env)
> plot(dune.Manure) 
 ----------- FAILURE REPORT -------------- 
 --- failure: length > 1 in coercion to logical ---
 --- srcref --- 
: 
 --- package (from environment) --- 
vegan
 --- call from context --- 
plot.cca(dune.Manure)
 --- call from argument --- 
!is.null(g$centroids) && !is.na(g$centroids)
 --- R stacktrace ---
where 1: plot.cca(dune.Manure)
where 2: NextMethod("plot", x, ...)
where 3: plot.rda(dune.Manure)
where 4: plot(dune.Manure)

 --- value of length: 10 type: logical ---
        RDA1 RDA2
Manure0 TRUE TRUE
Manure1 TRUE TRUE
Manure2 TRUE TRUE
Manure3 TRUE TRUE
Manure4 TRUE TRUE

[psolymos]

I couldn't trigger the error, but it looks like we have is.na(g$centroids)[1] at one place and is.na(g$centroids) at 3 more places in plot.cca. The report suggests that the second case (or all cases for that matter) should be any(is.na(g$centroids)) (or all(is.na(g$centroids)) when it is negated) to avoid having length >1 in if condition.

[eduardszoecs]

Can you set options(warn=2, error=recover) to turn the warnings into errors & start the debugger at warning/error?

[jarioksa]

@Edild the "problem" is that the function does not even give a warning in Linux or in Mac, but triggers an error in Windows. We also have a continuing checking in github both in Linux and in Windows, but these pass. There is something in the win-builder that I cannot reproduce, and mere setting of _R_CHECK_LENGTH_1_CONDITION_ does not help.

If I look at the internal code, I see that in example(cca) and there in plot(dune.Manure) the g$centroids returns length > 1:

> is.na(g$centroids)
         RDA1  RDA2
Manure0 FALSE FALSE
Manure1 FALSE FALSE
Manure2 FALSE FALSE
Manure3 FALSE FALSE
Manure4 FALSE FALSE

However, in Mac (and obviously in Linux, but I have no Linux at home) the whole condition reduces to length 1:

> length(g$centroids) > 0 && is.na(g$centroids)
[1] FALSE

It seems that this is not reduced to length 1 in win-builder or the checking happens in some wilder way.

@psolymos : the fix is indeed simple, like you said. It suffices to wrap the condition within all() or any() or explicitly indexing with [1L], but the problem is finding these places when I don't get an error. Of course I can go on with win-builder checks, but it sounds wasteful: the first error came from adonis, the second from cca and there are still help pages cca.object to wcmdscale to go in alphabetic order.

[jarioksa]

@Edild @psolymos I think I have solved the problem of testing, but cannot test this at home: R-devel has a new environment variable (https://cran.r-project.org/doc/manuals/r-devel/R-ints.html):

_R_CHECK_LENGTH_1_LOGIC2_
Optionally check if either argument of the binary operators && and || has length greater than one. The format is the same as for _R_CHECK_LENGTH_1_CONDITION_. Default: unset (nothing is reported)

This seems not be ported to release versions, and I don't have R-devel in my tiny laptop so that setting this variable does nothing. R-devel NEWS gives this as an "experimental" feature, but that does not mean that CRAN incoming check & the CRAN team would be forgiving here.

The length > 1 conditions has triggered a warning for ages, and we have fixed all those cases. So the remaining cases are just violations or _R_CHECK_LENGTH_1_LOGIC2_. If any of you has R-devel and a plenty of free time, you are free to find these violating pieces of code and fix them (@gavinsimpson ?).

I retired yesterday so that my access to a Linux computer w/R-devel at the uni is more limited now.

[jarioksa]

I said I wouldn't, but I built R-devel in my Macbook Air. This time I was more persistent and chased the libraries needed for building R. With this, I could find and fix the _R_CHECK_LENGTH_1_LOGIC2_ cases, and now vegan 2.5-4 passes win-builder tests. (Most fixes were obvious, but in ordixyplot the error was deep down in the grid package that provides the infrastructure for lattice, and I was lost for a long time).

This means that I can submit 2.5-4 on Sunday Feb 3 or the following Monday leaving some buffer for the Feb 8 deadline.

[legendre]

Dear Jari, You are truly incredible! Splendid work! I arrived in Barcelona this morning to attend the SIBECOL conference Monday to Thursday. http://www.congresosociedadibericaecologia2019.net/Conference_PROGRAM_496_p.htm <http://www.congresosociedadibericaecologia2019.net/Conference_PROGRAM_496_p.htm> Miquel De Caceres, who is a "local" here, picked me up at the airport. He will be presenting his new "Trajectory analysis" (freshly published last Thursday in Ecological Monographs) at the conference Tuesday at noon. It was good to see him again! Will will stroll through old Barcelona together tomorrow. Are you also in Spain? Will you be coming to the Conference? Lots of interesting people will be there :-) Pierre ========

…

Le 2 févr. 2019 à 19:53, Jari Oksanen ***@***.***> a écrit : I said I wouldn't, but I built R-devel in my Macbook Air. This time I was more persistent and chased the libraries needed for building R. With this, I could find and fix the _R_CHECK_LENGTH_1_LOGIC2_ cases, and now vegan 2.5-4 passes win-builder tests. (Most fixes were obvious, but in ordixyplot the error was deep down in the grid package that provides the infrastructure for lattice, and I was lost for a long time). This means that I can submit 2.5-4 on Sunday Feb 3 or the following Monday leaving some buffer for the Feb 8 deadline. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#305 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABf7nD-fxkvJSuy18A8YunXGdOXdjP6sks5vJd6qgaJpZM4aUGU3>.

--- Pierre Legendre ------------------- Courriel: Pierre.Legendre@UMontreal.CA Département de sciences biologiques, Université de Montréal, C.P. 6128, succ. Centre-ville, Tel.: (514) 343-7591 Montréal, Québec H3C 3J7, Canada Fax : (514) 343-2293 http://www.numericalecology.com/ ** Nouveaux livres -- ** Legendre & Legendre 2012: Numerical ecology, 3rd English ed (Elsevier). ** Borcard, Gillet & Legendre 2014: Numerical ecology with R, Chinese ed. ** Borcard, Gillet & Legendre 2018: Numerical ecology with R, 2nd English ed.

[jarioksa]

vegan 2.5-4 was released today with these fixes.