[R] transition from depends to imports

[Andreas Leha]

Duncan Murdoch <murdoch.duncan at gmail.com> writes:

> On 09/05/2014, 5:39 PM, Andreas Leha wrote:
>> Hi Duncan,
>>
>> Thank you for your follow-up and fast response!
>>
>> Duncan Murdoch <murdoch.duncan at gmail.com> writes:
>>
>>> On 09/05/2014, 5:08 PM, Andreas Leha wrote:
>>>> Hi all,
>>>>
>>>> I am having trouble to do a particular transition from depends to
>>>> imports in one of my packages.
>>>>
>>>> This packages uses 'wilcoxsign_test' from the 'coin' package.  And this
>>>> is the only function from the coin package that it uses (directly).
>>>>
>>>> (Everything works fine, as long as my package depends on the coin
>>>> package.)
>>>>
>>>> My guess was, that I would need to only import that function from the
>>>> coin package. (I.e. to move the coin package from Depends
>>>> to Imports in the DESCRIPTION and in the NAMESPACE do
>>>> 'importFrom(coin,wilcoxsign_test)'.)
>>>>
>>>> But then, calling wilcoxsign_test leads to this error:
>>>> ,----
>>>> | Error in formula2data(formula, data, subset, frame = parent.frame(), ...) :
>>>> |   could not find function "ModelEnvFormula"
>>>> `----
>>>>
>>>> Now, formula2data is an unexported function in the coin package and
>>>> ModelEnvFormula is a (exported) function in modeltools.
>>>>
>>>> So, I tried to import ModelEnvFormula from modeltools -- no avail.
>>>> And I tried to import formula2data from coin -- not possible
>>>> And I tried to import the whole coin package -- no avail.
>>>>
>>>> So, here is the question:  How do I import 'wilcoxsign_test' from the
>>>> coin package in a way, that it is usable?
>>>>
>>>
>>> Which version of the packages are you using?  The current version of
>>> coin imports ModelEnvFormula from modeltools, so it should be able to
>>> find that function.
>>>
>>> Duncan Murdoch
>>
>> As it seems, the version I use(d) depends on modeltools.
>>
>> ,----
>> | > packageDescription("coin")
>> | Package: coin
>> | Title: Conditional Inference Procedures in a Permutation Test Framework
>> | Date: 2013-04-26
>> | Version: 1.0-22
>> | Author: Torsten Hothorn, Kurt Hornik, Mark A. van de Wiel and Achim
>> |         Zeileis
>> | Maintainer: Torsten Hothorn <Torsten.Hothorn at R-project.org>
>> | Description: Conditional inference procedures for the general
>> |         independence problem including two-sample, K-sample
>> |         (non-parametric ANOVA), correlation, censored, ordered and
>> |         multivariate problems.
>> | Depends: R (>= 2.2.0), methods, survival, mvtnorm (>= 0.8-0),
>> |         modeltools (>= 0.2-9)
>> | Suggests: multcomp, xtable, e1071, vcd
>> | Enhances: Biobase
>> | LazyLoad: yes
>> | LazyData: yes
>> | License: GPL-2
>> | Packaged: 2013-04-26 12:33:15 UTC; hothorn
>> | NeedsCompilation: yes
>> | Repository: CRAN
>> | Date/Publication: 2013-04-26 15:27:04
>> | Built: R 3.0.1; x86_64-pc-linux-gnu; 2013-05-29 16:07:42 UTC; unix
>> |
>> | -- File: /usr/lib/R/site-library/coin/Meta/package.rds
>> `----
>>
>> I can confirm that updating coin (to version 1.1-0) did resolve my
>> problem.  Thank you very much!
>>
>>
>> This leads to the follow-up question:  How would I have correctly dealt
>> with this situation if my dependency ('coin') had not been updated?
>
> I think the only workaround would be for you to say your package
> Depends on modeltools.  It's really a coin issue, not yours, and the
> right people fixed it.
>

They did, indeed.  Thank you for the info and for your quick help on
this!

Regards,
Andreas