Dear list,
In using the scatter.smooth() function (modreg) on a small data set (100
obs) the following error was produced:
> scatter.smooth(Na, S)
Warning message:
k-d tree limited by memory. ncmax= 200
I haven't used scatter.smooth much but when I have, I haven't seen this
message before.
gc() returns
> gc()
used (Mb) gc trigger (Mb)
Ncells 417693 11.2 667722 17.9
Vcells 103949 0.8 786432 6.0
This is on a Win XP box with 512 MB RAM, with plenty of space still
available (Task Manager Reports 216 MB of RAM and pagefile in use).
I have looked through the help files for loess and loess.control but
didn't see anything about ncmax. Can someone explain why I might be
getting this warning?
Many thanks,
Gavin
> version
_
platform i386-pc-mingw32
arch i386
os mingw32
system i386, mingw32
status
major 1
minor 7.1
year 2003
month 06
day 16
language R
I can send the data if required.
--
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
Gavin Simpson [T] +44 (0)20 7679 5522
ENSIS Research Fellow [F] +44 (0)20 7679 7565
ENSIS Ltd. & ECRC [E] gavin.simpson at ucl.ac.uk
UCL Department of Geography [W] http://www.ucl.ac.uk/~ucfagls/cv/
26 Bedford Way [W] http://www.ucl.ac.uk/~ucfagls/
London. WC1H 0AP.
%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%
On Tue, 1 Jul 2003, Gavin Simpson wrote:> Dear list, > > In using the scatter.smooth() function (modreg) on a small data set (100 > obs) the following error was produced: > > > scatter.smooth(Na, S) > Warning message: > k-d tree limited by memory. ncmax= 200 > > I haven't used scatter.smooth much but when I have, I haven't seen this > message before. > > gc() returns > > > gc() > used (Mb) gc trigger (Mb) > Ncells 417693 11.2 667722 17.9 > Vcells 103949 0.8 786432 6.0 > > This is on a Win XP box with 512 MB RAM, with plenty of space still > available (Task Manager Reports 216 MB of RAM and pagefile in use). > > I have looked through the help files for loess and loess.control but > didn't see anything about ncmax. Can someone explain why I might be > getting this warning?It's trying to fit a very complicated function, and running out of its own internal storage, which is set in the C routine loess_workspace. However, 200 cells should be ample for 100 obs, so I guess the algorithm is not working properly in your example. -- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595