>
> Scott,
>
> I'm not much of a C programmer; I do know something about AUTOGRAPH.
>
> A call to "c_ezmxy" results in a call to "EZMXY", which, directly or
> indirectly, still calls the AGCHxx routines. Therefore, I think you
> can still use the FORTRAN versions of these routines. If you want to
> use C versions of them instead, you will have to give them modified
> names, the nature of which depends on the particular system you're
> using. The NCAR Graphics macro "NGCALLF" will give you the name you
> need, but I don't know just how to use that macro from your end; I'm
> familiar with using it as an NCAR Graphics developer.
>
> Dave Kennison
Scott,
Dave is correct, a call to "c_ezmxy" eventually results in a call
to the AGCHxx routines.
Here's a C example that shows how to create a C version of AGCHNL.
This example is equivalent to the Fortran AGEX09 example. It calls
"c_ezxy" instead of "c_ezmxy", but the important thing is how agchnl
is implemented:
#include <stdio.h>
#include <math.h>
#include <ncarg/ncargC.h>
#define max(x,y) ((x) > (y) ? (x) : (y) )
#define min(x,y) ((x) < (y) ? (x) : (y) )
#define pow2(x) ((x)*(x))
#define WSTYPE 8
#define WKID 1
main()
{
/*
* Define the data arrays.
*/
int i, idim, xdim, ydim, mlow, mhgh;
float xdat[400],ydat1[200],ydat[400], xmax, dlow, dhgh,power;
float ymin, ymax;
extern void bndary();
/*
* Initialize gks.
*/
gopen_gks("stdout",0);
gopen_ws(WKID, NULL, WSTYPE);
gactivate_ws(WKID);
/*
* Fill the data arrays.
*/
for( i = 0; i < 400; i++ ) {
xdat[i]=((float)(i+1)-1.)/399.;
}
xdim = 200;
ydim = 1;
idim = 200;
mlow = 3;
mhgh = 3;
dlow = .01;
dhgh = 10.;
(void)gendat(ydat1,idim,xdim,ydim,mlow,mhgh,dlow,dhgh);
for( i = 0; i < 200; i++ ) ydat[i] = ydat1[i];
dlow = -10.;
dhgh = -.01;
(void)gendat(ydat1,idim,xdim,ydim,mlow,mhgh,dlow,dhgh);
for( i = 200; i < 400; i++ ) ydat[i] = ydat1[i-200];
/*
* The y data ranges over both positive and negative values.
* it is desired that both ranges be represented on the same
* graph and that each be shown logarithmically, ignoring
* values in the range -.01 to +.01, in which we have no
* interest. First we map each y datum into its absolute
* value (.01 if the absolute value is too small). then we
* take the base-10 logarithm, add 2.0001 (so as to be sure
* of getting a positive number), and re-attach the original
* sign. We can plot the resulting y data on a linear y axis.
*/
for( i = 0; i < 400; i++ ) {
xmax = fabs((double)ydat[i]) > .01 ? fabs((double)ydat[i]) : .01;
power = (float)log10((double)xmax)+2.0001;
if( ydat[i] < 0 ) ydat[i] = -fabs((double)power);
else ydat[i] = fabs((double)power);
if( i ) {
ymin = ydat[i] < ymin ? ydat[i] : ymin;
ymax = ydat[i] > ymax ? ydat[i] : ymax;
}
else {
ymin = ymax= ydat[0];
}
}
/*
* In order that the labels on the y axis should show the
* original values of the y data, we change the user-system-
* to-label-system mapping on both y axes and force major
* ticks to be spaced logarithmically in the
* label system (which will be defined by the subroutine
* agutol in such a way as to re-create numbers in the
* original range).
*/
c_agseti("LEFT/FUNCTION.",1);
c_agseti("LEFT/MAJOR/TYPE.",2);
c_agseti("RIGHT/FUNCTION.",1);
c_agseti("RIGHT/MAJOR/TYPE.",2);
/*
* Change the left-axis label to reflect what's going on.
*/
c_agsetc("LABEL/NAME.","L");
c_agseti("LINE/NUMBER.",100);
c_agsetc("LINE/TEXT.","LOG SCALING, POSITIVE AND NEGATIVE$");
/*
* Draw a boundary around the edge of the plotter frame.
*/
bndary();
/*
* Draw the curve.
*/
c_ezxy(xdat,ydat,400,"EXAMPLE 9$");
/*
* Close gks.
*/
gdeactivate_ws(WKID);
gclose_ws(WKID);
gclose_gks();
}
gendat (data,idim,m,n,mlow,mhgh,dlow,dhgh)
float *data, dlow, dhgh;
int idim, m, n, mlow, mhgh;
{
/*
* This is a routine to generate test data for two-dimensional graphics
* routines. Given an array "DATA", dimensioned "IDIM x 1", it fills
* the sub-array ((DATA(I,J),I=1,M),J=1,N) with a two-dimensional field
* of data having approximately "MLOW" lows and "MHGH" highs, a minimum
* value of exactly "DLOW" and a maximum value of exactly "DHGH".
*
* "MLOW" and "MHGH" are each forced to be greater than or equal to 1
* and less than or equal to 25.
*
* The function used is a sum of exponentials.
*/
float ccnt[3][50], fovm, fovn, dmin, dmax, temp;
extern float fran();
int nlow, nhgh, ncnt, i, j, k, ii;
fovm=9./(float)m;
fovn=9./(float)n;
nlow=max(1,min(25,mlow));
nhgh=max(1,min(25,mhgh));
ncnt=nlow+nhgh;
for( k=1; k <= ncnt; k++ ) {
ccnt[0][k-1]=1.+((float)m-1.)*fran();
ccnt[1][k-1]=1.+((float)n-1.)*fran();
if (k <= nlow) {
ccnt[2][k-1]= -1.;
}
else {
ccnt[2][k-1] = 1.;
}
}
dmin = 1.e36;
dmax = -1.e36;
for( j = 1; j <= n; j++ ) {
for( i = 1; i <= m; i++ ) {
ii = (i-1)*n+j-1;
data[ii]=.5*(dlow+dhgh);
for( k = 1; k <= ncnt; k++ ) {
temp = -(pow2((fovm*((float)(i)-ccnt[0][k-1])))+
pow2(fovn*((float)(j)-ccnt[1][k-1])));
if (temp >= -20.) data[ii]=data[ii]+.5*(dhgh-dlow)
*ccnt[2][k-1]*exp(temp);
}
dmin=min(dmin,data[ii]);
dmax=max(dmax,data[ii]);
}
}
for( j = 0; j < m*n; j++ ) {
data[j]=(data[j]-dmin)/(dmax-dmin)*(dhgh-dlow)+dlow;
}
return(1);
}
void bndary()
{
/*
* Draw a line showing where the edge of the plotter frame is.
*/
c_plotif (0.,0.,0);
c_plotif (1.,0.,1);
c_plotif (1.,1.,1);
c_plotif (0.,1.,1);
c_plotif (0.,0.,1);
c_plotif (0.,0.,2);
}
float fran()
{
/*
* Pseudo-random-number generator.
*/
static double x = 2.718281828459045;
extern double fmod();
x = fmod(9821.*x+.211327,1.);
return((float)x);
}
NGCALLF(agutol,AGUTOL)(iaxs,funs,idma,vinp,votp)
int *iaxs,*idma;
float *funs,*vinp,*votp;
{
/*
* Left or right axis.
*/
if ( *funs == 1.) {
if ( *idma < 0) {
if( *vinp < 0 ) {
*votp = -fabs(log10(max(fabs(*vinp),.01))+2.0001);
}
else {
*votp = fabs(log10(max(fabs(*vinp),.01))+2.0001);
}
}
else {
if( *vinp < 0 ) {
*votp = -fabs(pow(10.,(fabs(*vinp)-2.0001)));
}
else {
*votp = fabs(pow(10.,(fabs( *vinp)-2.0001)));
}
}
}
/*
* All others.
*/
else {
*votp = *vinp;
}
return(1);
}
NGCALLF(agchnl,AGCHNL)(iaxs,vils,chrm,mcim,ncim,ipxm,chre,mcie,ncie)
char *chrm, *chre;
int *iaxs, *mcim, *ncim, *ipxm, *mcie, *ncie;
float *vils;
{
/*
* Modify the left-axis numeric label marking the value "0.".
*/
if ( *iaxs == 1 && *vils == 0.) {
chrm[0] = ' ';
*ncim=1;
*ipxm=0;
*ncie=0;
}
return(1);
}
> > From ncarg-talk-admin AT unknown Tue Sep 10 10:56:00 2002
> > To: ncarg-talk AT unknown
> > Subject: C routines for autograph
> >
> > Hi,
> > I am trying to update a fortran program that uses autograph to run in C and
> > I am having problems with the AGCHCU, AGCHAX, AGCHNL subroutines. These are
> > dummy routines that ezmxy calls that the user can alter to change line
> > color and other parameters. Does the c_ezmxy routine call similar dummy
> > functions ???
> > I'd rather not rewrite all the code from scratch (everything works except
> > these routines).
> > Any help would be appreciated.
> > Thanks.
> > ______________________
> >
> > Scott L. Goodrick, Ph.D.
> > Research Meteorologist
> > USDA Forest Service
> > 320 Green Street
> > Athens, GA 30602
> > Phone: 706- 559-4237
> _______________________________________________
> ncarg-talk mailing list
> ncarg-talk AT unknown
> http://mailman.ucar.edu/mailman/listinfo/ncarg-talk
>
_______________________________________________
ncarg-talk mailing list
ncarg-talk AT unknown
http://mailman.ucar.edu/mailman/listinfo/ncarg-talk
This archive was generated by hypermail 2b29 : Tue Sep 10 2002 - 17:25:27 MDT