void conmatr (x, y, z, npts, xgmin, xgmax, nxpts, ygmin,  
         ygmax, nypts, zorig, zstep, zmax, lnstyle, labint) 

void conmatrNS (x, y, z, npts, xgmin, xgmax, nxpts, ygmin,  
         ygmax, nypts, zorig, zstep, zmax, lnstyle, labint)

CONMATR.C Level 3
 
 float *x, *y, *z The x, y, and z data arrays
 int npts The dimension of the  x, y, and z data arrays.
float xgmin The minimum x grid value.
float xgmax The maximum x grid value.
int nxpts The number of x grid intersections.
float ygmin The minimum y grid value.
float ygmax The maximum y grid value.
int nypts The number of y grid intersections.
float zorig The minimum z value.
float zstep The size of the z step.
float zmax The maximum z value.
int *lnstyle An array of coded numbers for changing line styles and/or colors for each contour level. See the explanation at the beginning of this section. This specification will not handle the extended GraphiC colors and is retained for compatibility purposes. 
If lnstyle == NULL, the more powerful specification capability of contcolor() will be used.
int labint Labels are placed on every labint contour. If labint = 0, contour labels are suppressed.
 
conmatr() and conmatrNS() produce contour plots from irregularly spaced data points. Remember that good results depend on having data points that provide good coverage of the plotting region. They each call InterpolateZ() to get interpolated data and then call the contour plotter to produce the plot. conmatr() first divides the grid by 4 to produce a coarse grid, then calls InterpolateZ(). The interpolated data is then smoothed using a cubic spline. conmatrNS() provides no smoothing. conmatr() automates the method used in the example R3TEST.C to draw a contour plot. See InterpolateZ()

If you get the message "Too few data points, try a coarser grid ," try reducing nxpts and nypts. If you cannot get satisfactory results that way, try decreasing the precision in R3PLT.C. This can be done by calling settolerance(). The default value for tolerance is 3.0. The larger the number, the less the precision.