Re: [ferret_users] spectrum and FFTA error

[Ansley Manke <ansley.b.manke@xxxxxxxx>]

Hi,
It sounds as if you have solved the problems.  I would just like to add that the error message,
   Time axis must be a regular axis

does not have to do with missing data, but with Ferret having detected the time axis as irregularly-spaced.  The command

yes? show grid variable_name  ! or

yes? show axis axis_name

lists the name, number of points, and some letters: "r" or "i" for regular or irregular spacing, "m" if the axis is "modulo" (repeating) coordinates, and a - sign if the coordinates (on a Z axis) are decreasing downward.  So the Taxis here, is regular.

yes? use monthly_navy_winds
yes? show grid uwnd
    GRID GDN1
 name       axis              # pts   start                end
 FNOCX     LONGITUDE          144mr   20E                  17.5E(377.5)
 FNOCY     LATITUDE            73 r   90S                  90N
 normal    Z
 TIME      TIME               132 r   16-JAN-1982 20:00    17-DEC-1992 03:30
 normal    E
 normal    F

but this monthly dataset, with coordinates defined to represent exact calendar months, has an irregularly-spaced time axis.

yes? use "http://www.esrl.noaa.gov/psd/thredds/dodsC/Datasets/ncep.marine/sst.mean.nc"
yes? sh grid sst
    GRID GCK1
 name       axis              # pts   start                end
 LON       LONGITUDE          180mr   1E                   1W
 LAT       LATITUDE            90 r   89S                  89N
 normal    Z
 TIME      TIME               242 i   01-JAN-1991 00:00    01-FEB-2011 00:00
 normal    E
 normal    F

Sometimes Ferret sees the time axis as irregular, but it should be treated as regularly spaced. This can happen depending on how the coordinates were written to the file. If this is the case and if it's a NetCDF or OPeNDAP dataset, you can open the data with

yes? use/regulart dataset.nc

-Ansley

On 11/10/2013 11:31 PM, chinnu sachi wrote:

Thank you very much for a fast reply. I set path like as you mentioned, in bashrc and now go gmt_power is working fine.

On Mon, Nov 11, 2013 at 12:25 PM, Akshay Hegde <akshay.k.hegde@xxxxxxxxx> wrote:

Reinstallation won't help. After installation you need to set path for GMT executables.

if you are Ubuntu user and installed through either sudo apt-get install gmt or through Ubuntu Software Center or through Synaptic Package Manager then default path would be  /usr/lib/gmt/bin set this in your .bashrc or in profile otherwise if you installed manually by downloading tar or zip then set respective path of GMT.

Thanks and Regards,
Akshay Hegde 
CSIR-National Institute of Oceanography,
Dona Paula, Panjim, Goa -403004

 

On Mon, Nov 11, 2013 at 12:19 PM, chinnu sachi <chinnu.sachi7@xxxxxxxxx> wrote:

Yes, Iam getting an error when i tried the command spawn spectrum1d.

So as u said,the problem is with the GMT environment, I reinstalled GMT now, still it is showing same error.

On Mon, Nov 11, 2013 at 12:13 PM, Akshay Hegde <akshay.k.hegde@xxxxxxxxx> wrote:

look at the difference here... I suspect there might be some problem with your GMT Environment Settings (path/bash profile etc) if its not working even with default ferret datasets.

Here spectrum1d is working

yes? spawn spectrum1d
spectrum1d 4.5.6 - Compute auto- [and cross- ] spectra from one [or two] timeseries

usage:  spectrum1d -S<segment_size> [-C[<xycnpago>]] [-D<dt>] [-H[i][<nrec>]] [-N<name_stem>]
    [-V] [-W] [-b[i|o][s|S|d|D[<ncol>]|c[<var1>/...]]] [-f[i|o]<colinfo>]

Here I just tried some test command "spectrum1d_test_command" since it does not exist so getting following error

yes? spawn spectrum1d_test_command
sh: 1: spectrum1d_test_command: not found

Thanks and Regards,
Akshay Hegde 



CSIR-National Institute of Oceanography,
Dona Paula, Panjim, Goa -403004

On Mon, Nov 11, 2013 at 12:02 PM, Akshay Hegde <akshay.k.hegde@xxxxxxxxx> wrote:

Did you try with default dataset which comes with ferret ? whether following commands are working ?

    yes? SET DATA monthly_navy_winds
    yes? GO gmt_power uwnd[x=180,y=0] uwnd x180y0 64
    yes? PLOT x180y0_power
    yes? GO overlay_bars x180y0_power  x180y0_error red

Thanks and Regards,
Akshay Hegde 
CSIR-National Institute of Oceanography,
Dona Paula, Panjim, Goa -403004

On Mon, Nov 11, 2013 at 11:55 AM, chinnu sachi <chinnu.sachi7@xxxxxxxxx> wrote:

My time axis is ok,I list it below. even I removed clobber ,still same error is coming.

 I can open spectrum1d with man spectrum1d

         VARIABLE : L
                        axis TIME
             FILENAME : v_test.nc
             SUBSET   : 600 points (TIME)
 15-JAN-1958 00 /   1:    1.0
 15-FEB-1958 00 /   2:    2.0
 15-MAR-1958 00 /   3:    3.0
 15-APR-1958 00 /   4:    4.0
 15-MAY-1958 00 /   5:    5.0
 15-JUN-1958 00 /   6:    6.0
 15-JUL-1958 00 /   7:    7.0
 15-AUG-1958 00 /   8:    8.0
 15-SEP-1958 00 /   9:    9.0
 15-OCT-1958 00 /  10:   10.0
 15-NOV-1958 00 /  11:   11.0
 15-DEC-1958 00 /  12:   12.0
 15-JAN-1959 00 /  13:   13.0
 15-FEB-1959 00 /  14:   14.0
 15-MAR-1959 00 /  15:   15.0
 15-APR-1959 00 /  16:   16.0
 15-MAY-1959 00 /  17:   17.0
 15-JUN-1959 00 /  18:   18.0
 15-JUL-1959 00 /  19:   19.0
 15-AUG-1959 00 /  20:   20.0
 15-SEP-1959 00 /  21:   21.0
 15-OCT-1959 00 /  22:   22.0
 15-NOV-1959 00 /  23:   23.0
 15-DEC-1959 00 /  24:   24.0

On Mon, Nov 11, 2013 at 11:48 AM, Akshay Hegde <akshay.k.hegde@xxxxxxxxx> wrote:

Sorry I forgot to mention make sure that you can open man page using following command since you are getting this error "sh: 1: spectrum1d: not found"

man spectrum1d

Thanks and Regards,
Akshay Hegde 
CSIR-National Institute of Oceanography,
Dona Paula, Panjim, Goa -403004

On Mon, Nov 11, 2013 at 11:40 AM, Akshay Hegde <akshay.k.hegde@xxxxxxxxx> wrote:

Try something like this

GO gmt_power variable[x=180,y=0] grid output seg_len

and in your script you used both clobber and append

append ---> If file exists then append, otherwise create and write to it

clobber ---> If the file exists then overwrite it, otherwise create and write to it

in your case you need to use only append

and Hows your Time axis ?

list t[gt=variable]

Thanks and Regards,
Akshay Hegde 
CSIR-National Institute of Oceanography,








Dona Paula, Panjim, Goa -403004

On Mon, Nov 11, 2013 at 10:32 AM, chinnu sachi <chinnu.sachi7@xxxxxxxxx> wrote:

Dear ferret users,

                   I am getting an error error message while trying to run the gmt_power function. The details of the error for power spectrum as follows:

yes? go power.jnl
set mem/size = 650
 Cached data cleared from memory
use v_test.nc
 repeat/k=1:`20` ( go gmt_power V_NEW GHA1 gmt 256; list/file=test/clobber/append gmt_power;plot gmt_power)
 !-> repeat/k=1:20 ( go gmt_power V_NEW GHA1 gmt 256; list/file=test/clobber/append gmt_power;plot gmt_power)
!-> REPEAT: K=1
 LISTing to file gmt.tseries
>>> Using GMT command: spectrum1d gmt.tseries -Ngmt -S256 -D30.43583
sh: 1: spectrum1d: not found
 **TMAP ERR: non-existent or not on line
             gmt.xpower
SET DATA/EZ/var=freq gmt.xpower
Command file, command group, or REPEAT execution aborted

And also when using FFTA command, error is coming like "Bailing out of external function "ffta":
     Time axis must be a regular axis
 **ERROR: error in external function" .But when gone through the mail archives i found a link stating that it happens when there is missing values. But in my data there is no missing value, in that region.

Thanks in advance

--
Regards,
Chinnu Sachidanandan
Physical Oceanography Division
National Institute of Oceanography
csachidanandan@xxxxxxx

--
Regards,
Chinnu Sachidanandan
Physical Oceanography Division
National Institute of Oceanography
csachidanandan@xxxxxxx

--
Regards,
Chinnu Sachidanandan
Physical Oceanography Division
National Institute of Oceanography
csachidanandan@xxxxxxx

--
Regards,
Chinnu Sachidanandan
Physical Oceanography Division
National Institute of Oceanography
csachidanandan@xxxxxxx