[Rcpp-devel] Package built with Rcpp fails to install

[Simon Zehnder]

Tal,

looking at the first error I see on the Linux Machine: 

cut_lower.cpp: In function 'void find_dend_for_height(Rcpp::List, std::vector<Rcpp::RObject_Impl<Rcpp::PreserveStorage> >&, double)':
cut_lower.cpp:35:22: error: could not convert 'tree' from 'Rcpp::List {aka Rcpp::Vector<19>}' to 'Rcpp::RObject'
    if(get_height(tree) <= height) {

It says that the Rcpp::List object cannot be simply converted to an RObject - which is needed in the get_height function. I do not know the inheritance structure between RObject and List. What I would try though is to use the Rcpp::as<Rcpp::RObject>(list) as parameter in get_height. Another possibility is maybe to use instead of an RObject and RObject* pointer as a parameter. 

Best

Simon




On 03 Feb 2014, at 22:29, Tal Galili <tal.galili at gmail.com> wrote:

> Just to correct my previous statement - 
> It seems that I after I restarted RStudio, the package no longer compiles for me.
> I tried win-builder and it also failed, see:
> http://win-builder.r-project.org/ziuFii9ZaY9t/00install.out
> 
> So the good news is that I can work on this. The bad news is that I still have no idea why is this happening or what to try/read in order to resolve it.
> 
> So if/when someone in this list would be able to help give me a direction, I would gladly try to follow it through.
> 
> Best,
> Tal
> 
> 
> 
> 
> 
> ----------------Contact Details:-------------------------------------------------------
> Contact me: Tal.Galili at gmail.com |  
> Read me: www.talgalili.com (Hebrew) | www.biostatistics.co.il (Hebrew) | www.r-statistics.com (English)
> ----------------------------------------------------------------------------------------------
> 
> 
> 
> On Mon, Feb 3, 2014 at 11:06 PM, Tal Galili <tal.galili at gmail.com> wrote:
> Dear all,
> 
> Romain - thanks. Your input is valuable, whenever you can spare it.
> 
> Kevin - You are spot on. I was indeed using 0.10.6 (although, to my defense, it is the most updated Windows version on CRAN at the moment...).
> Also, I now got 0.11.0 from github, and I was still able to compile my package on my machine without any error (e.g: with Rcpp 0.11.0 on Windows). So I am back to square one of not being able to reproduce the error on my machine.
> 
> Simon - thanks for checking this also.
> 
> 
> 
> With regards,
> Tal
> 
> 
> 
> 
> ----------------Contact Details:-------------------------------------------------------
> Contact me: Tal.Galili at gmail.com |  
> Read me: www.talgalili.com (Hebrew) | www.biostatistics.co.il (Hebrew) | www.r-statistics.com (English)
> ----------------------------------------------------------------------------------------------
> 
> 
> 
> On Mon, Feb 3, 2014 at 10:44 PM, Simon Zehnder <szehnder at uni-bonn.de> wrote:
> Dirk,
> 
> I saw this message shortly after I sent my second one …
> 
> Simon
> On 03 Feb 2014, at 21:33, Dirk Eddelbuettel <edd at debian.org> wrote:
> 
> >
> > On 3 February 2014 at 21:24, Simon Zehnder wrote:
> > | Hi all,
> > |
> > | compiled the package from github on a Linux Scientific Machine:
> >
> > I just did too, under Ubuntu 13.10 at work.  Works fine for me on that
> > machine but with Rcpp 0.10.6.
> >
> > However, as of this morning, that is no longer the CRAN version.  I spent
> > three weeks testing these things and logged most of the changes in the
> > rcpp-logs repo with the RcppCore account on GitHub. Kevin tested all of
> > BioC.  Such compilation failures affected maybe half a dozen packages,
> > probably less.  More needed Imports: / NAMESPACE changes, and I'll follow up
> > on this.
> >
> > Let's calm this thread down for now.  We upgraded Rcpp, and Tal drew the
> > shorter stick and now needs to update his package.
> >
> > Dirk
> >
> > --
> > Dirk Eddelbuettel | edd at debian.org | http://dirk.eddelbuettel.com
> 
> 
> 
> _______________________________________________
> Rcpp-devel mailing list
> Rcpp-devel at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel