If a 3D plot is ok you might want to try the persp() command. You just give the
thing your matrix with each entry corresponding to the height (z) at point (x,
y). It's pretty straightforward to use and can handle a lot of extra options.
M
Zitat von Clayton Nall <nall(a)fas.harvard.edu>du>:
I am trying to produce a contour or shaded contour
plot similar to those we
produced in Gov 2001. However, unlike the examples we dealt with in Gov
2001, where we were evaluating likelihood functions, the set of points that
I'm using does not fall neatly on a grid. Instead, I have an n x 3 table,
each row of which contains a simulated observation point (x, y, z) that I
would like to represent graphically.
Can anyone recommend a good canned function that would allow me to produce
smooth graphs from data of this type? The functions contour() and image()
seem limited to evaluating functions over a grid. The function symbols()
offers flexibility in terms of where points are located but doesn't produce
smoothed graphs.
Thanks,
Clayton
--
Martin Kanz
Ph.D. Student
Political Economy & Government
Harvard University
Cambridge, MA 02138
U.S.A.