Tweet
The original posting...
I am presently implementing GCVSPL for Matlab 6.0 (Windows NT), with
the source files downloaded from the ISB site (Tony Reina's versions).
I have modified the mat_files from those used for version 5.1 due to
the apparent change in memory management by Matlab as outlined in the
Matlab help files. However, this solution seems to be unstable and
erratic. While the code will run multiple times in Matlab (sometimes),
it will often crash out when I attempt to plot any of the smoothed
data. In addition, the code will only run once when called in Matlab
from within LabVIEW.
My assessment is that I may have problems with my gateway modifications!
Has anyone run across a similar issue when implementing the GCVSPL-code
in Matlab 6.0 or earlier? Alternatively, has anyone implemented Woltring's
algorithm in LabVIEW?
************************************************** **************************
Thanks to Ty Hedrick, Trey Crisco and Tony Reina for their replies.
Ty's build of Tony's code has been running on NT4 and Matlab 6.0 for him
with no problems. Interestingly, it worked the first time for me, but
thereafter gave me the same memory errors. (A strange southern hemisphere
effect perhaps, although Trey reported similar Matlab problems compiling
other code).
The smoothing solution: I only wanted to use a quintic spline (e.g. Reinsch,
1967), and the university had a license for the spline toolbox...."spaps"
works a treat!
I remain curious about the memory problem, and any solutions or suggestions
are most welcome.
Rachel Skoss
Department of Human Movement and Exercise Science
The University of Western Australia
Crawley, WA, 6009, Australia
rskoss@cyllene.uwa.edu.au
Ph: +61 8 9380 7355
Fax: +61 8 9380 1039
---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------
I am presently implementing GCVSPL for Matlab 6.0 (Windows NT), with
the source files downloaded from the ISB site (Tony Reina's versions).
I have modified the mat_files from those used for version 5.1 due to
the apparent change in memory management by Matlab as outlined in the
Matlab help files. However, this solution seems to be unstable and
erratic. While the code will run multiple times in Matlab (sometimes),
it will often crash out when I attempt to plot any of the smoothed
data. In addition, the code will only run once when called in Matlab
from within LabVIEW.
My assessment is that I may have problems with my gateway modifications!
Has anyone run across a similar issue when implementing the GCVSPL-code
in Matlab 6.0 or earlier? Alternatively, has anyone implemented Woltring's
algorithm in LabVIEW?
************************************************** **************************
Thanks to Ty Hedrick, Trey Crisco and Tony Reina for their replies.
Ty's build of Tony's code has been running on NT4 and Matlab 6.0 for him
with no problems. Interestingly, it worked the first time for me, but
thereafter gave me the same memory errors. (A strange southern hemisphere
effect perhaps, although Trey reported similar Matlab problems compiling
other code).
The smoothing solution: I only wanted to use a quintic spline (e.g. Reinsch,
1967), and the university had a license for the spline toolbox...."spaps"
works a treat!
I remain curious about the memory problem, and any solutions or suggestions
are most welcome.
Rachel Skoss
Department of Human Movement and Exercise Science
The University of Western Australia
Crawley, WA, 6009, Australia
rskoss@cyllene.uwa.edu.au
Ph: +61 8 9380 7355
Fax: +61 8 9380 1039
---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------