Hi Mary,
Julie Arblaster has just pointed out to me that setting 'jave' to an even
value was causing the problem. With an odd value for jave, the script works
ok now. Thanks for your interest.
-- Harun=================================================================== Dr. M. Harun Ar Rashid Bureau of Meteorology Research Centre Email: h.rashid@bom.gov.au PO Box 1289K Tel : +61 3 9669 4550 Melbourne, Victoria 3001 Fax : +61 3 9669 4660 Australia http://www.bom.gov.au/bmrc ===================================================================
-----Original Message----- From: Mary Haley [mailto:haley@ucar.edu] Sent: Wednesday, 4 May 2005 14:03 To: Harun Rashid Cc: Dennis Shea Subject: Re: Problem using 'specx_anal' repeatedly
Hi Harun,
Can you provide me with a script and data so I can try running it here? If the data file is too big, you can ftp it ftp.cgd.ucar.edu:
ftp ftp.cgd.ucar.edu anonymous email cd incoming put ... put ... quit
You'll have to send me the names of the files, because I can't do an "ls" on the directory.
Thanks,
--Mary
On Wed, 4 May 2005, Harun Rashid wrote:
> Hi, > > I was trying to calculate temporal power spectra at a number of longitude > points. When using the 'specx_anal' > function in a do-loop, it causes a Segmentation fault, i.e, could not do the > following: > > Rspec = new( (/ntimh,nlon/), "float") > do i = 0, nlon-1 > RtD = dtrend(Rt(:,i),False) ; detrend data > spec = specx_anal(RtD,iopt,jave,pct) ; calculate spectrum > Rspec(:,i) = spec@spcx > end do > > Where, Rt is Rt(ntim,nlon) and ntimh = ntim/2. (Using specx_anal only once > works fine). I've tried all sorts of things to avoid the above problem, but > with no results (the problem still persists if I delete the last line: > Rspec(:,i) = spec@spcx). Does anyone have any suggestion how to get around > this problem? Thanks. > > -- > Harun > > =================================================================== > Dr. M. Harun Ar Rashid > Bureau of Meteorology Research Centre Email: h.rashid@bom.gov.au > PO Box 1289K Tel : +61 3 > 9669 4550 > Melbourne, Victoria 3001 Fax : +61 3 9669 > 4660 > Australia > http://www.bom.gov.au/bmrc > =================================================================== > > > _______________________________________________ > ncl-talk mailing list > ncl-talk@ucar.edu > http://mailman.ucar.edu/mailman/listinfo/ncl-talk >
This archive was generated by hypermail 2b29 : Wed May 04 2005 - 08:07:14 MDT