[SciPy-User] Scipy's Delaunay triangulation

barbara.padova barbara.padova@gmail....
Tue Feb 21 08:26:11 CST 2012


I am noticing an unexplained behaviour when use scipy's (0.9.0)
Delaunay triangulation routine. My points are UTM coordinates stored
in numpy.array([[easting, northing], [easting, northing], [easting,
northing]]).
If I use the true coordinates Scipy's edges are missing some of my
points.
If I calculate the mean_easting and the mean_northing values and
change my array in numpy.array([[easting-mean_easting , northing-
mean_northing ], [easting-mean_easting , northing-mean_northing ],
[easting-mean_easting , northing-mean_northing ]]), the result it's
OK.
I know that there are  rounding errors and I also tried using the data
type float64, but whit my original date in UTM I have the same
problems.
There is a solution to use my original data with scipy's Delaunay
triangulation routine?

Thanks for any help!
Best regards,
Barbara Padova


More information about the SciPy-User mailing list