similar to: lbfgsb from C/C++

Full_Name: Rod Montgomery Version: 1.1.0 OS: Windows-95 Submission from: (NULL) (38.26.56.3) File R-1.1.0/src/appl/lbfgsb.c -- Comment in code says two papers describing underlying math are in ftp://ece.nwu.edu/pub/lbfgs/lbfgs_bcm/ but that directory seems not to exist. (Actually, the comment does not give a URL - it just says to look in that directory on that FTP server.) Paper [1] does

Hi. I just noticed the paper by Morales and Nocedal Remark on "Algorithm 778: L-BFGS-B: Fortran Subroutines for Large-Scale Bound Constrained Optimization". TOMS 2011; 38(1): 7 http://www.ece.northwestern.edu/~morales/PSfiles/acm-remark.pdf which describes a couple of improvements (speed and accuracy) to the original Netlib code which AFAICT is that still used by optim() via f2c.

Hi, Mark et al.: Thanks, Mark. Three comments: 1. Rvmmin was one of the methods I tried after Ravi directed me to optimx. It returned NAs for essentially everything. See my email of this subject stamped 4:43 PM Central time = 21:43 UTC. 2. It would be interesting to know if the current algorithm behind optim and optimx with

Can anyone help me with lbfgsb (function for optimization)? It takes the following parameters: void lbfgsb (int n, int lmm, double *x, double *lower, double *upper, int *nbd, double *Fmin, optimfn fn, optimgr gr, int *fail, void *ex, double factr, double pgtol, int *fncount, int *grcount, int maxit, char *msg, int trace, int nREPORT); What do I put for parameter ex (11th parameter)? I looked at

Dear List, I stepped into a strange effect which I can't explain to myself (probably due to lack of knowledge on R internals). I have four vectors a,b,c and z of size 10000 each. With these vectors I call .Call("hyp2f1forrey", a, b, b, z, PACKAGE = "hyp2f1") to access SEXP hyp2f1forrey(SEXP a, SEXP b, SEXP c, SEXP x) { int i,n; double *xa, *xb, *xc, *xx,

Hi! I want to return a matrix. The code does the R interfacing. This version does it fine. SEXP ans,dim; PROTECT(ans = NEW_NUMERIC(count*2)); memcpy(NUMERIC_POINTER(ans),result,count*sizeof(double)); memcpy(&(NUMERIC_POINTER(ans)[count]),occur,count*sizeof(double)); /** PROTECT(dim=NEW_INTEGER(2)); INTEGER_POINTER(dim)[0]=2; INTEGER_POINTER(dim)[1]=count;

Hi when I call the function below in R, i get the error: Object 'pairlist' can't be converted to 'double'. #include <R.h> #include <Rdefines.h> #include <Rmath.h> SEXP CSimPoisson(SEXP lambda, SEXP tgrid, SEXP T2M, SEXP Ni, SEXP NT) { double sign, EVar; double *xlambda, *xtgrid, *xT2M, *xNi, *xNT, *xtau; SEXP tau; int ltgrid =0; int i = 0; int j = 0;

Hi, I wanted to use R's lbfgsb method for minimization from C. Unfortunately, my toy examples always crashes (segmentation fault). What's wrong with it? double eval(int n, double* par, void *ex) { double result = 0; for (int i=0; i<n; ++i) { result += par[i]*par[i]; } printf("result=%.2f\n", result); return result; } void grad(int n, double *par, double *gr,

Hi all, im triyng to implement a bayesian model with R and c++. I have a strange problem. I can't reproduce the error with a small script and then i post the original one. The problem is after the line for(MCMC_iter2=0;MCMC_iter2<thin;MCMC_iter2++) For the first 34 iterations all work fine, after, all the simulations of mu_acc_P return an "nan". If i delete the line

Hi I have a FORTRAN version of the L-BFGS-B algorithm and I was comparing it to the code in the lbfgsb.c file available at R-4.0.0.tar.gz Everithing looks the same, except for those two lines that must be printed by the prn3lb function in case of an error (lines 3559 and 3561 in lbfgsb.c): case -5: Rprintf("l(%d) > u(%d). No feasible solution", k, k); break; case -7:

Hello, I would like, if it is possible, to compare the effect of a variable across regression models. I have looked around but I haven't found anything. Maybe someone could help? Here is the problem: I am studying the effect of a variable (age) on an outcome (local recurrence: lr). I have built 3 models: - model 1: lr ~ age y = \beta_(a1).age - model 2: lr ~ age + presentation

Greetings, I am trying to call simple C-code from R. I am on Windows XP with RTools installed. The C-function is #include <R.h> #include <Rinternals.h> #include <Rmath.h> #include <Rdefines.h> // prevent name mangling extern "C" { SEXP __cdecl test(SEXP s){ SEXP result; PROTECT(result = NEW_NUMERIC(1)); double* ptr=NUMERIC_POINTER(result); double t =

Full_Name: Geordon Marchak Version: R-1.3.1 OS: Solaris 2.6 Submission from: (NULL) (199.67.239.91) Got the following error when compiling on Solaris 2.6 with the standard GNU gcc package (from the Sun site). BTW - linux compiles no problem (and fast too :-) # make `Makedeps' is up to date. gcc -I. -I../../src/include -I../../src/include -I/usr/local/include -DHAVE_CONFIG_H -g -O2 -c

Full_Name: Chris Pelton Version: R-1.5.0 OS: Solaris & Submission from: (NULL) (169.237.121.210) I am unable to compile R-1.5.0, and am getting the following errors with make. I searched through the buglist and didn't see anything similar, although I wouldn't doubt if I was just using the wrong version of something. gcc version 3.0.3 gcc -I. -I../../src/include

Greetings, I am trying to call simple C-code from R. I am on Windows XP with RTools installed. The C-function is #include <R.h> #include <Rinternals.h> #include <Rmath.h> #include <Rdefines.h> // prevent name mangling extern "C" { SEXP __cdecl test(SEXP s){ SEXP result; PROTECT(result = NEW_NUMERIC(1)); double* ptr=NUMERIC_POINTER(result); double t =

Hello, It appears that the file ./appl/lbfgsb.c defines setulb() as a static function, and it is included in optim.c, so setulb() is not exported by the R library. I have some sofware that uses setulb() directly, and I would prefer to avoid having to recompile this sofware to export that function. Is there any way to get access to setulb(), or must I use the lower and upper params in optimix?

>>>>> Spencer Graves <spencer.graves at prodsyse.com> >>>>> on Sat, 8 Oct 2016 18:03:43 -0500 writes: [.............] > 2. It would be interesting to know if the > current algorithm behind optim and optimx with > method='L-BFGS-B' incorporates Morales and Nocedal (2011) > 'Remark on ?Algorithm 778:

Hi I have a FORTRAN version of the L-BFGS-B algorithm and I was comparing it to the code in the lbfgsb.c file available at R-4.0.0.tar.gz Everithing looks the same, except for those two lines that must be printed by the prn3lb function in case of an error (lines 3559 and 3561 in lbfgsb.c): case -5: Rprintf("l(%d) > u(%d). No feasible solution", k, k); break; case -7:

I want to write a little stand-alone C program that calls R_alloc, but I get a segmentation fault: int main(int argc, char** argv){ double* d = (double *)R_alloc(2, sizeof(double)); // <- segmentation fault! return 0; } gdb reveals that sizeof(double) evaluated to 0: > R_alloc (nelem=2, eltsize=0) at memory.c:1649 so it results in the segfault later. This is how I compile my

I'm continuing my work on finding speedups in generalized inverse calculations in some simulations. It leads me back to .C and .Call, and some questions I've never been able to answer for myself. It may be I can push some calculations to LAPACK in or C BLAS, that's why I realized again I don't understand the call by reference or value semantics of .Call Why aren't users of