[Met_help] [rt.rap.ucar.edu #53284] History for Problem in Point-Stat for GFS vs GTS
John Halley Gotway via RT
met_help at ucar.edu
Tue Feb 7 11:10:02 MST 2012
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Dear Met-Help,
I'm trying to run point-stat for GFS gridded data (fcst) in NetCDF format
output from Rscript and GTS as obs data (output from ascii2nc). It is
giving some error as follows:
ERROR: str_to_prob_info() -> problems parsing the string "*,*".
I have tried many combinations to define variable in PointStatConfig_rf
file and each time I'm getting same kind of problem. Why it is happening
that's I don't know?
Please help me to sort out this problem.
PS: All the data files and script files have been transferred in
incoming/irap/harvir_data
Thanx in advance.
--
* Regards*
*********************************
* HARVIR SINGH <http://www.harvir007.webs.com/>*
* Project Scientist**
HCL @ NCMRWF
***********************************
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #53284] Problem in Point-Stat for GFS vs GTS
From: John Halley Gotway
Time: Mon Feb 06 09:59:48 2012
Harvir,
OK, I see the problem. I wasn't able to fully test the fix because it
looks like there was a problem in how you posted the data to the ftp
site. The contents of the files named
"GTS_OBS_RRF_2011-06-25-000-26-29_76-79.nc" and "PointStatConfig_rf"
are identical. It looks like you accidentally overwrote the NetCDF
point observation file.
But hopefully the following information will get you up and running.
I see the following settings in the PointStatConfig file you sent:
fcst_field[] = [ "APCP_24(*,*)" ];
obs_field[] = [ "APCP_06(*,*)" ];
Since you're verifying a NetCDF forecast file, the setting for the
fcst_field[] works fine. However, you're using the same type of
setting for the observation field, and that's where the problem is.
You need to specify the verifying observation field using the GRIB
code conventions. So please try using the following settings:
fcst_field[] = [ "APCP_24(*,*)" ];
obs_field[] = [ "APCP/A24" ];
That tells Point-Stat to compare the forecast NetCDF field
"APCP_24(*,*)" against any observations with GRIB code = 61 (which is
APCP) and an accumulation interval of 24 hours.
Please let us know if you continue to experience problems.
Thanks,
John Halley Gotway
On 02/06/2012 12:22 AM, Harvir Singh via RT wrote:
>
> Mon Feb 06 00:22:09 2012: Request 53284 was acted upon.
> Transaction: Ticket created by harvirduphy at gmail.com
> Queue: met_help
> Subject: Problem in Point-Stat for GFS vs GTS
> Owner: Nobody
> Requestors: harvirduphy at gmail.com
> Status: new
> Ticket<URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=53284>
>
>
> Dear Met-Help,
>
>
> I'm trying to run point-stat for GFS gridded data (fcst) in NetCDF
format
> output from Rscript and GTS as obs data (output from ascii2nc). It
is
> giving some error as follows:
> ERROR: str_to_prob_info() -> problems parsing the string "*,*".
>
> I have tried many combinations to define variable in
PointStatConfig_rf
> file and each time I'm getting same kind of problem. Why it is
happening
> that's I don't know?
> Please help me to sort out this problem.
>
> PS: All the data files and script files have been transferred in
> incoming/irap/harvir_data
>
> Thanx in advance.
------------------------------------------------
Subject: Problem in Point-Stat for GFS vs GTS
From: Harvir Singh
Time: Mon Feb 06 21:56:56 2012
Thank you John for your prompt reply. Now its working as per your
suggestion.
Than you again.
On 6 February 2012 22:29, John Halley Gotway via RT
<met_help at ucar.edu>wrote:
> Harvir,
>
> OK, I see the problem. I wasn't able to fully test the fix because
it
> looks like there was a problem in how you posted the data to the ftp
site.
> The contents of the files named
> "GTS_OBS_RRF_2011-06-25-000-26-29_76-79.nc" and "PointStatConfig_rf"
are
> identical. It looks like you accidentally overwrote the NetCDF
point
> observation file.
>
> But hopefully the following information will get you up and running.
I
> see the following settings in the PointStatConfig file you sent:
>
> fcst_field[] = [ "APCP_24(*,*)" ];
> obs_field[] = [ "APCP_06(*,*)" ];
>
> Since you're verifying a NetCDF forecast file, the setting for the
> fcst_field[] works fine. However, you're using the same type of
setting
> for the observation field, and that's where the problem is.
> You need to specify the verifying observation field using the GRIB
code
> conventions. So please try using the following settings:
>
> fcst_field[] = [ "APCP_24(*,*)" ];
> obs_field[] = [ "APCP/A24" ];
>
> That tells Point-Stat to compare the forecast NetCDF field
"APCP_24(*,*)"
> against any observations with GRIB code = 61 (which is APCP) and an
> accumulation interval of 24 hours.
>
> Please let us know if you continue to experience problems.
>
> Thanks,
> John Halley Gotway
>
>
> On 02/06/2012 12:22 AM, Harvir Singh via RT wrote:
> >
> > Mon Feb 06 00:22:09 2012: Request 53284 was acted upon.
> > Transaction: Ticket created by harvirduphy at gmail.com
> > Queue: met_help
> > Subject: Problem in Point-Stat for GFS vs GTS
> > Owner: Nobody
> > Requestors: harvirduphy at gmail.com
> > Status: new
> > Ticket<URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=53284>
> >
> >
> > Dear Met-Help,
> >
> >
> > I'm trying to run point-stat for GFS gridded data (fcst) in NetCDF
format
> > output from Rscript and GTS as obs data (output from ascii2nc). It
is
> > giving some error as follows:
> > ERROR: str_to_prob_info() -> problems parsing the string "*,*".
> >
> > I have tried many combinations to define variable in
PointStatConfig_rf
> > file and each time I'm getting same kind of problem. Why it is
happening
> > that's I don't know?
> > Please help me to sort out this problem.
> >
> > PS: All the data files and script files have been transferred in
> > incoming/irap/harvir_data
> >
> > Thanx in advance.
>
>
--
* Regards*
*********************************
* HARVIR SINGH <http://www.harvir007.webs.com/>*
* Project Scientist**
HCL @ NCMRWF
***********************************
------------------------------------------------
More information about the Met_help
mailing list