I didn't get any feedback and i'm posting the message again.I hope i'm not
bothering any of you but please if you have any ideas about the problem i'm
facing in v.surf.rst please let me now because i must get over it as quickly as
possibly.
So as i said before my region is :
And just before the end of the v.surf.rst calculations
i'm getting the message G realloc : out of memory and then the procedure stops
I'working on a pentium 4 cpu 3.00ghz with 2gb of ram.
note that a segmentation algorithm is used when an extreme number of input
points are present- which by the way how many are there, and how evenly are
they distributed across your region?
Cheers,
Dylan
On Thursday 10 August 2006 00:35, Harris Pappas wrote:
hello,
I didn't get any feedback and i'm posting the message again.I hope i'm not
bothering any of you but please if you have any ideas about the problem i'm
facing in v.surf.rst please let me now because i must get over it as
quickly as possibly.
So as i said before my region is :
And just before the end of the v.surf.rst calculations
i'm getting the message G realloc : out of memory and then the procedure
stops I'working on a pentium 4 cpu 3.00ghz with 2gb of ram.
Thanks for the response. The number of points of my vector file (contour lines)
is 7730470
3642623 is used by v.surf.rst.
Could you explain my about the segmax value i didn't get it much ? As for the
dmax. I've read in the manual the info about this value. As i put the
resolution (25) in g.region the value changed automaticaly. Is that ok? or
might causing problems ? For now i'm trying to see what will happen with the CV
procedure but it takes very long as expected. About the segmentation procedure
is being discribed in the manual, can i do anything about it? i guess no if i
get it right. As you can see i'm new to GRASSGIS so sorry if i make silly
questions.
I don't recommend using rst algorithm for interpolating on irregular,
sparse input as the contour lines usually are. Please browse the list
archives as this issue pops up from time to time.
For contour lines and orther clustered input datasets I personally find
natural neighbor algorithm most reliable. On Grass WIKI you'll find
r.surf.nnbathy script which is a Grass wrapper for nnbathy nat. neigh.
interp. library.
On Thursday 10 August 2006 12:48, Harris Pappas wrote:
Dylan,
Thanks for the response. The number of points of my vector file (contour
lines) is 7730470
3642623 is used by v.surf.rst.
Could you explain my about the segmax value i didn't get it much ? As for
the dmax. I've read in the manual the info about this value. As i put the
resolution (25) in g.region the value changed automaticaly. Is that ok? or
might causing problems ? For now i'm trying to see what will happen with
the CV procedure but it takes very long as expected. About the segmentation
procedure is being discribed in the manual, can i do anything about it? i
guess no if i get it right. As you can see i'm new to GRASSGIS so sorry if
i make silly questions.