Skip to content

[Rcpp-devel] RcppArmadillo Mat_meat.hpp:4467 and arma_static_check.hpp:31 errors

5 messages · Patrick Veenstra, Dirk Eddelbuettel

Original
Patrick Veenstra
# 
When attempting to compile a cpp function that uses RcppArmadillo (in
RStudio), I get the following two errors from files provided by
RcppArmadillo:

Error 1: In ~/RLibs/RcppArmadillo/include/armadillo_bits/Mat_meat.hpp line
4467:

no matching function for call to ?arma::Gen<arma::Mat<double>,
arma::gen_zeros>::apply(arma::Mat<unsigned int>&) const?

Error 2: in
~/RLibs/RcppArmadillo/include/armadillo_bits/arma_static_check.hpp line 31:

size of array is negative and uninitialized const 'junk'

Environment details:

Platform: Linux fedora core 22, 4.0.6-300.fc22.x86_64
Armadillo version: 6.700.6
R: 3.3.0
RStudio: 1.0.143

NAMESPACE Contents:

exportPattern("^[[:alpha:]]+")
importFrom(Rcpp, evalCpp)
useDynLib(ThisPackage)

DESCRIPTION links to both Rcpp and RcppArmadillo

Any idea on how I could go about solving this?

Many thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rcpp-devel/attachments/20170724/eb47bd00/attachment.html>
Patrick Veenstra
# 
Correction, Armadillo version: 7.900.2
On 24 July 2017 at 19:09, Patrick Veenstra <patrick at pveenstra.com> wrote:

            

      
      
      
    

        
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rcpp-devel/attachments/20170724/eb8fc295/attachment.html>

  
  


  


      

    

    

      
      

        
      

    

    

      
      

        


  

        

      Patrick Veenstra
    

    

      
      #
    

  


  

      
My apologies! I mistook what the code in those files were for. It seemed to
be completely separate from my code. So I thought it was something about my
environment. But in the process of setting up a reproducible example, I
found that actually it was something in my own code. An accidental
assignment of zeros<mat>(x,y) to a umat. Odd that this error came through
in those files rather than directly in my cpp file. Anyway, lesson learnt!
Don't assume anything!

Thanks

        
On 24 July 2017 at 23:40, Dirk Eddelbuettel <edd at debian.org> wrote:

        

            

      
      
      
    

        
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/rcpp-devel/attachments/20170725/d44e7ef7/attachment.html>

  
  


  


      

    

    

      
      

        


  

        

      Dirk Eddelbuettel
    

    

      
      #
    

  


  

      
On 25 July 2017 at 12:56, Patrick Veenstra wrote:

        
| My apologies! I mistook what the code in those files were for. It seemed to
| be completely separate from my code. So I thought it was something about my
| environment. But in the process of setting up a reproducible example, I
| found that actually it was something in my own code. An accidental
| assignment of zeros<mat>(x,y) to a umat. Odd that this error came through
| in those files rather than directly in my cpp file. Anyway, lesson learnt!
| Don't assume anything!

Yes, been there, done that -- that is one of many reasons why creating a
reproducible example is a good exercise.  I too had numerous report
dispappear that way before sending :)

Dirk