abed@saturn.wustl.edu (Abed M. Hammoud) (03/16/91)
Hello, I have a small question, Its an easy one, so I hope I will not get flamed... A friend of mine wants to plot a set of Polar coordinates data in 3D. The data is in the form: r0, phi0, z0 r1, phi1, z1 . . . . . . rn, phin, zn I wrote a routine for him that takes this data set and do the obvious polar to rectangular transformation: x=round(r*cos(phi)*nbin+nbin+1) y=round(r*sin(phi)*nbin+nbin+1) X(x,y)=z The results I get is acceptable, but since I have to round x and y to integer numbers I loose some accuracy, and because there is a small (not large enough) number of polar coordinates samples the 3D function I get is not smooth. Also, I have to pass to the routine the number of samples that correspond to maximum radius, (nbin). nbin determines how big the rectangular array X will be. So, my question is; is this the best way of dealing with the problem ?. My friend seems to think that there must be a better way. He checked Auto-Cadd and there they seem to have a routine that do the same job but when you call the function you don't have to pass an equivalent of nbin, which my routine needs. Notes: (Auto-cadd is $3000 and he cannot afford to buy it). (The routine I wrote uses matlab). (I can interpolate to make the function smoother, that will make the 2D array even bigger..) Any suggestion are appreciated.... Thanks.... -------------------------------------------------------------- | Abed M. Hammoud abed@saturn.wustl.edu| | Washington University. office:(314)726-7547 | | Electronic Systems & Signals Research Laboratory. | | Dept. of Electrical/Biomedical Engineering. | | St.Louis Mo U.S.A | --------------------------------------------------------------