[R-pkg-devel] Checking should be performed on sources prepared by 'R CMD build'?

I'm building a new package and got a warning I have not seen before. This is probably simple, but I don't understand what the warning means. I'm using RCMD check from the command line and get the warning: "Checking should be performed on sources prepared by 'R CMD build'." The full

Re: [Rd] Problems with building a private R-package with Fortran source under Windows

On Windows, there is a separate file makevars.win (not makevars) that is used at compile/link time. See the R-exts manual, particularly section 5.8.2. Hope that helps, John .. John P. Nolan Math/Stat Department 227 Gray Hall,

Re: [Rd] historical significance of Pr(Chisq) 2.2e-16

Presumably from .Machine$double.eps [1] 2.220446e-16 Whether this means the tail probability is actually that small, or that the routine that computes it can't get any more accuracy than that, you'll have to dig deeper. John .. John

Re: [Rd] Win 64 package build - ERROR: loading failed for 'x64'

Ripley Sent by: r-devel-boun...@r-project.org Date: 06/24/2012 01:38AM Subject: Re: [Rd] Win 64 package build - ERROR: loading failed for 'x64' On 24/06/2012 02:54, John Nolan wrote: I have developed an R package that works under Win32, but when I attempt to build it on Win64, I get ERROR

[Rd] Win 64 package build - ERROR: loading failed for 'x64'

I have developed an R package that works under Win32, but when I attempt to build it on Win64, I get ERROR: loading failed for 'x64' More precisely, I developed and tested the package under Win32 and it works. But when I move to a 64 bit Windows 7 (Home Premium) system, and attempt to build

Re: [Rd] 0.5 != integrate(dnorm,0,20000) = 0

I have wrestled with this problem before. I think correcting the warning to absolute error ~= 0 is a good idea, and printing a warning if subdivisions==1 is also helpful. Also, including a simple example like the one that started this thread on the help page for integrate might make the issue

Re: [Rd] 0.5 != integrate(dnorm,0,20000) = 0

and pick an appropriate region of integration. John Nolan, American U. -r-devel-boun...@r-project.org wrote: - To: r-devel@r-project.org From: Pierre Chausse Sent by: r-devel-boun...@r-project.org Date: 12/07/2010 09:46AM Subject: Re: [Rd] 0.5 != integrate(dnorm,0,2) = 0 The warning

Re: [Rd] Suggested change to integrate.Rd (was: Re: 0.5 != integrate(dnorm, 0, 20000) = 0)

sense. John Nolan, American U -Spencer Graves spencer.gra...@structuremonitoring.com wrote: - To: John Nolan jpno...@american.edu From: Spencer Graves spencer.gra...@structuremonitoring.com Date: 12/07/2010 07:58PM Cc: pchau...@uwaterloo.ca, r-devel@r-project.org Subject: Suggested

[Rd] constrOptim( ): conflict between help page and code

There is a contradiction between what the help page says and what constrOptim actually does with the constraints. The issue is what happens on the boundary. The help page says The feasible region is defined by ‘ui %*% theta - ci = 0’, but the R code for constrOptim reads if (any(ui %*%

[Rd] cascade option for dev.new?

,y,type='l',main='second plot') John Nolan Math/Stat Dept. American University __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel

Re: [Rd] Probability of data values form empirical distribution

The function ecdf(x) computes the empirical cdf from data in the vector x. You can plot it with plot(ecdf(x)), or compute the emp. cdf at new values, e.g. my.cdf - ecdf(x) my.cdf( 0:3 ) # computes the emp. cdf at 0,1,2,3 John

Re: [Rd] Inaccurate complex arithmetic of R (Matlab is accurate)

Ravi, There has been a lot of chatter about this, and people don't seem to be reading carefully. Perhaps this will help clarify things. The problem appears to be that R was evaluating x^2 not as multiplication x*x, but as x^2.0=exp(2.0*log(x)), using standard C functions for the complex log

Re: [Rd] Can't get F77_CALL(dgemm) to work [SEC=UNCLASSIFIED]

Daniel, R apparently uses Fortran order AND storage method when storing a matrix. For an (n x m) matrix, Fortran allocates a single block of nm doubles and stores them in the order A(1,1),A(2,1),A(3,1),...,A(n,1),A(1,2),A(2,2),...,A(n,m). In contrast, C allocates a vector of n pointers, each

Re: [Rd] invalid comparison in numeric sequence (PR#13551)

This is one of the joys of floating point arithmetic. Look at the 7th element of seq(0,1,.1) - 0.6 Also, check the all.equal(x,y,tolerance=epsilon ) function. John -r-devel-boun...@r-project.org wrote: - To: r-de...@stat.math.ethz.ch From: alexandre.court...@gmail.com Sent by:

Re: [Rd] sources of code; was Generate random numbers in Fortran

I think something like this would be very useful. I'm sure you know of NIST's Guide to Mathematical Software (GAMS). It looks like they list places to find things, but it is clear some of that is proprietary (IMSL, NAG), some is on NETLIB but seems to have embedded copyright statements. (I once

[Rd] Rtools28 - undefined references with gfortran

): undefined reference to `_gfortran_concat_string Any guidance on how to solve this problem? John Nolan [[alternative HTML version deleted]] __ R-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-devel

[Rd] RCMD SHLIB: static libraries and f77 libraries on Windows

Dear R-devel, I am converting some stand-alone programs (mixed C and F77) to R functions. I've run into two issues that I haven't been able to resolve. I've looked at the R-exts manual and Readme files, but haven't found answers. (1) Can I link to a (Win32) static library? Is there some