| Author |
 Topic  |
|
|
ddecotiis
6 Posts |
 Posted - 09/01/2009 : 5:30:40 PM
|
Origin Ver. 8 and Service Release 6(Select Help-->About Origin):
Operating System: Xp
I'm trying to use the NAG x01ajc function via. the X01AJC constant in the code of a curve fitting routine. When I try to compile the code I get the following error:
"Error, Function x02ajc@LAAAAAAA was called, but its body could not be located during linking."
I presume that I need to add the .c or .dll file containing the function body to the workspace, but I'm not sure which file to add. Can anyone shed some light on the problem?
-Thanks
|
|
|
Iris_Bai
China
Posts |
Posted - 09/01/2009 : 10:33:19 PM
|
Please include header file and compile again.
#include <\NAG8\nagx02.h>
Iris |
 |
|
|
ddecotiis
6 Posts |
Posted - 09/01/2009 : 10:44:29 PM
|
Iris,
Thanks for the quick response. I'm getting a new crops of errors:
"C:\Program Files\OriginLab\Origin8\OriginC\system\NAG8\nagx02.h(113) :Error, illegal function argument type"
"C:\Program Files\OriginLab\Origin8\OriginC\system\NAG8\nagx02.h(113) :Error, syntax error in variable declaration."
"C:\Documents and Settings\Dan\My Documents\OriginLab\Origin8\User Files\OriginC\NLSF\_nlfLigand_Binding_MR_Solver.fit(4) :Error, include file compilation error"
Any other ideas?
Thanks again
|
|
|
|
Iris_Bai
China
Posts |
|
|
ddecotiis
6 Posts |
Posted - 09/02/2009 : 4:50:25 PM
|
Iris,
I've already got that include in my .c source file. Any other ideas?
Thanks |
|
|
|
Iris_Bai
China
Posts |
|
|
ddecotiis
6 Posts |
Posted - 09/16/2009 : 2:03:47 PM
|
Iris,
Sorry about the delay. Thanks again for agreeing to take a look at this. I've uploaded the function .FDF file to the originlab /incoming directory. The filename is Ligand_Binding_MR_Solver_.FDF. I'm still getting the same error on compilation. Let me know what you think.
Thanks again,
|
|
|
|
ddecotiis
6 Posts |
Posted - 09/16/2009 : 3:52:20 PM
|
Iris,
There was some issue with the code. I've uploaded a file with correct syntax to UPDATED_Ligand_Binding_MR_Solver_.FDF.
Thanks |
|
|
|
larry_lan
China
Posts |
Posted - 09/16/2009 : 11:42:51 PM
|
Hi:
We cannot reproduce your compiling error in UPDATED_Ligand_Binding_MR_Solver_.FDF. I wonder if you updated to SR6 from other patches? Say from SR5 to SR6? If you keep your licence, how about download the demo installation file and repair Origin with your licence? Please let us know whether it works after repair.
Thanks
Larry |
|
|
|
ddecotiis
6 Posts |
Posted - 09/22/2009 : 8:32:49 PM
|
All,
I think there is a bigger problem with my function.
1. Compiling the function in the demo version of origin still results in the following error:
Function x02ajc@LAAAAAAA was called, but its body could not be located during linking.
I wrote a smaller example copying the NAG documentation code to demonstrate my point. The sample test.c code (shown below) will generate this error if the comment line is removed.
2. Any call to the NAG function for finding the root of system of transcendental functions (thread safe), nag_zero_nonlin_eqns_1 (C05tbc) generates the following error "External call execution error". I reproduced the sample code from the NAG documentation as test.c below. I am unable to reproduce the program output shown in the NAG documentation because of the repeated error.
test.c/
#include <Origin.h>
#include <OC_nag8.h>
#define NMAX 9
static void f(Integer n, double x[], double fvec[], Integer *userFlag, Nag_User *comm)
{
Integer k;
for (k=0; k<n; ++k)
{
fvec[k]=(3.0-x[k]*2.0)*x[k]+1.0;
if (k>0)
fvec[k] -= x[k-1];
if (k<n-1)
fvec[k] -= x[k+1]*2.0;
}
}
main()
{
double ffvec[NMAX];
double fx[NMAX];
Integer i,j;
double xtol;
static NagError fail;
Nag_User comm;
Integer n= NMAX;
printf("Program Results:\n");
for (j=0;j<n;j++)
x[j]=-1.0;
//xtol= sqrt(X02AJC); Commented till I can fix this problem
xtol = 0.00001;
c05tbc(n,x,fvec,f,xtol,&comm,&fail);
if (fail.code == NE_NOERROR)
{
printf("Final Approximate Solutions\n\n");
for(j=0;j<n;j++)
printf("%12.4f%s",x[j], (j%3==2 || j==n-1) ? "\n" : " ");
}else
{
printf("error occured: %s", fail.message);
}
}
/test.c
Any input would be great
Thanks |
|
|
|
Iris_Bai
China
Posts |
Posted - 09/25/2009 : 02:02:26 AM
|
Hi,
Send email to your mailbox, please check.
Iris |
Edited by - Iris_Bai on 09/25/2009 02:20:46 AM |
|
|
| |
Topic |
|
Machine tolerance function in curve fitting routin