There's an R object that has the machine precision, which could be a
reasonable threshold.
.Machine$double.eps
I believe there is a similar constant in the C++ standard library.
You might also try checking the condition of the matrix instead of the
determinant, but you might take a performance h
Rcpp should make this very straightforward; assuming you are putting this
in a package (because why wouldn't you?) you would just need to set some
flags in src/Makefile so that the external library gets picked up correctly
by R, and write a few Rcpp wrapper functions for interacting with it.
See a
lso using the nlohmann/json.h, maybe I can ask
him to
move json.h from src/ to inst/include for his next release. I think that's
all that needed
to make Rcpp::depends / LinkingTo work.
On Thu, Oct 11, 2018 at 12:18 PM Dirk Eddelbuettel wrote:
>
> On 11 October 2018 at 11:50, Neal Fultz
to inst/include).
Maybe someone else will find this useful some day.
Best,
Neal Fultz
___
Rcpp-devel mailing list
Rcpp-devel@lists.r-forge.r-project.org
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rcpp-devel
I had ported the bayesm version a while a back, here it is:
List rwishart(int nu, NumericMatrix const& V){// function to draw
from Wishart (nu,V) and IW// // W ~ W(nu,V) E[W]=nuV// // WI=W^-1
E[WI]=V^-1/(nu-m-1)
RNGScope rngscope; int m = V.nrow();mat Vm(V.begin(), m, m,
false);// Can'
I had ported the bayesm version a while a back, here it is:
List rwishart(int nu, NumericMatrix const& V){// function to draw
from Wishart (nu,V) and IW// // W ~ W(nu,V) E[W]=nuV// // WI=W^-1
E[WI]=V^-1/(nu-m-1)
RNGScope rngscope; int m = V.nrow();mat Vm(V.begin(), m, m,
false);// Can'
I use the debian packages. Upgrading r-cran-rcpp from testing to unstable
seems to have fixed this for me, so good call.
On Fri, Jul 12, 2013 at 10:25:17PM -0400, Krzysztof Sakrejda wrote:
> On Fri, Jul 12, 2013 at 8:50 PM, Dirk Eddelbuettel wrote:
> >
> > On 12 July 2013 at 12
I've been updating the C in an r package to use Rcpp and started seeing
so odd results from a function that samples from a discrete
distribution.
I think I've narrowed it down to the sugar runif. The two programs below
are identical except f_works uses unif_rand and f_broke goes through runif.
