http://qs1969.pair.com?node_id=480545


in reply to Delaunay triangulation
in thread Millions of line segment intersection calcs: Looking for speed tips

I for one would agree you are right on both counts, but to paraphrase Shakespeare, 'Does a problem by any other name solve more sweetly?'

Update: To clarify what I mean: the OP knows the problem and its solution but is looking for the fastest way to achieve it with perl.

One world, one people

Replies are listed 'Best First'.
Re^2: Delaunay triangulation
by tmoertel (Chaplain) on Aug 03, 2005 at 16:36 UTC
[reply]
Re^2: Delaunay triangulation
by tmoertel (Chaplain) on Aug 03, 2005 at 16:57 UTC
    To clarify what I mean: the OP knows the problem and its solution but is looking for the fastest way to achieve it with perl.

    How do you know what the OP knows and is looking for? Do you have some knowledge beyond what is in the original post? The OP says he is "willing to entertain ideas of differing algorithms." What makes you think otherwise?

    Cheers,
    Tom

    Tom Moertel : Blog / Talks / CPAN / LectroTest / PXSL / Coffee / Movie Rating Decoder

[reply]
Re^2: Delaunay triangulation
by runrig (Abbot) on Aug 03, 2005 at 19:39 UTC
    I tend to side with tmoertel. I once in ages past implemented my own naive algorithm for this very thing ( O(n^3) or O(n^4), probably the same way that the OP is doing, but in AutoLISP using AutoCAD), and although I knew there must be a better way, I didn't know at the time that just knowing the name Delaunay Triangulation would've helped in finding better algorithms (and it was long before Google and probably before Al Gore invented the internet). I've since been unable to find the time to implement this in perl :-(

    There was a start made on this in perl utilizing a C library, but all I see is a readme file.

[reply]

In Section Seekers of Perl Wisdom