Re: limit for axis values?[B

[Ansley Manke <Ansley.B.Manke@xxxxxxxx>]

Hi Mulyono,
There is not a limit on the length of axes that you define explicitly
with a DEFINE AXIS command, as long as you have enough
memory to work with the variables you define.

I think you just have to change the syntax a little.  Each time you
do the save/append, the variable REC is just one time step, so
you need to say the following, adding to REC the variable zero
at [L=`L`], which has the effect of putting the variable REC
at time L:

> ...
> let REC=h1+h2+h3+h4+h5+h6;\
> let fft_inv_olr=REC+zero[L=`L`];\
> save/append/file= olr.5daily.rec6.40snxglobe.nc fft_inv_olr)
Also, possibly it was just a typo in your email, but you defined
fft_inv_olr, and then saved fft_inv.

Ansley Manke

Mulyono Prabowo wrote:

Hi ferreters,

I have been calculating inverse harmonic analyses mannually (I am using
ferret 5.51 on SGI). I used script as bellows:

use/regulart olr.5daily.fftrffti.40snxglobe.nc
let Pi=3.1415927
DEFINE AXIS/T=3-JAN-1979:29-DEC-2001/npoints=1679 tpentad
LET zero=T[gt=tpentad@mod] * 0
! ****** 6ath HARMONIC
REPEAT/l=1:1679 (\
let h1=fftr[k=1]*cos(2*Pi*`l`*1/73)+ffti[k=1]*sin(2*Pi*`l`*1/73);\
let h2=fftr[k=2]*cos(2*Pi*`l`*2/73)+ffti[k=2]*sin(2*Pi*`l`*2/73);\
let h3=fftr[k=3]*cos(2*Pi*`l`*3/73)+ffti[k=3]*sin(2*Pi*`l`*3/73);\
let h4=fftr[k=4]*cos(2*Pi*`l`*4/73)+ffti[k=4]*sin(2*Pi*`l`*4/73);\
let h5=fftr[k=5]*cos(2*Pi*`l`*5/73)+ffti[k=5]*sin(2*Pi*`l`*5/73);\
let h6=fftr[k=6]*cos(2*Pi*`l`*6/73)+ffti[k=6]*sin(2*Pi*`l`*6/73);\
let REC=h1+h2+h3+h4+h5+h6;let fft_inv_olr=REC+zero;\
save/append/file= olr.5daily.rec6.40snxglobe.nc fft_inv)

When I ran this script, it stoped at 6th harmonic with warning as follows:

!-> REPEAT: L=1
 !-> DEFINE VARIABLE h1=fftr[k=1]*cos(2*Pi*1*1/73)+ffti[k=1]*sin(2*Pi*1*1/73)
 !-> DEFINE VARIABLE h2=fftr[k=2]*cos(2*Pi*1*2/73)+ffti[k=2]*sin(2*Pi*1*2/73)
 !-> DEFINE VARIABLE h3=fftr[k=3]*cos(2*Pi*1*3/73)+ffti[k=3]*sin(2*Pi*1*3/73)
 !-> DEFINE VARIABLE h4=fftr[k=4]*cos(2*Pi*1*4/73)+ffti[k=4]*sin(2*Pi*1*4/73)
 !-> DEFINE VARIABLE h5=fftr[k=5]*cos(2*Pi*1*5/73)+ffti[k=5]*sin(2*Pi*1*5/73)
 !-> DEFINE VARIABLE h6=fftr[k=6]*cos(2*Pi*1*6/73)+ffti[k=6]*sin(2*Pi*1*6/73)
 LISTing to file olr.5daily.rec6.40snxglobe.nc
 **TMAP ERR: error in line definition
             disordered output coordinate value: 683448 Axis: TPENTAD
LIST/FORMAT=CDF/append/file= olr.5daily.rec6.40snxglobe.nc fft_inv
Command file, command group, or REPEAT execution aborted

It ran well for 1st harmonic (REC = h1, only) until 5th harmonic (REC =
h1+h2+h3+h4+h5). It stoped when start calculating 6th harmonic. Does any
one have any idea what was wrong. Is there limit for axis value?

Thank you very much for your help.
regards,

Mulyono

--
Ansley Manke  Pacific Marine Environmental Laboratory  Seattle WA