[Met_help] [rt.rap.ucar.edu #86419] History for missing Orank when calculate rhist with STAT_analysis
John Halley Gotway via RT
met_help at ucar.edu
Wed Aug 8 13:35:36 MDT 2018
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Zhang, Mandy has shared OneDrive for Business files with you. To view them, click the links below.
<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw>
[https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw>
ensemble_stat_ensmr01_20180510_070000V.stat<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw>
<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A>
[https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A>
EnsmStatConfig_APCP24<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A>
<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ>
[https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ>
met_stat_analysis_mon_acc_phist.sh<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ>
Hi All,
I want to use Stat_analysis to get seasonal mean rank histogram, I find even if I set ORANK = true in ensemble_stat configure file, the ORANK can't be found in ensemble**.stat file, and it causes the error information when I run stat_analysis mode:
WARNING: do_job_aggr_stat() -> no matching STAT lines found for job: -job aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST -line_type ORANK -dump_row /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat -out_line_type PHIST -out_bin_size 0.05000
*****************************
But I have separated files as: ensemble_stat_ensmr01_20180510_060000V_orank.nc,
ensemble_stat_ensmr01_20180510_060000V_phist.txt,
ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
Do you know what's issue of my case?
I attached the ensemble_stat configure file, ensemble_stat**.stat and script of aggreate rhist with Stat_analysis.
Thank you very much!
Mandy
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: Julie Prestopnik
Time: Tue Jul 31 15:43:29 2018
Hi Mandy. Our MET expert is out of the office this week.
Unfortunately, I
am unable to help with this issue. I will have our MET expert follow
up
next week.
Julie
On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu>
wrote:
>
> Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> Transaction: Ticket created by Man.Zhang-1 at ou.edu
> Queue: met_help
> Subject: missing Orank when calculate rhist with STAT_analysis
> Owner: Nobody
> Requestors: Man.Zhang-1 at ou.edu
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
>
> Zhang, Mandy has shared OneDrive for Business files with you. To
view
> them, click the links below.
>
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> >
>
> ensemble_stat_ensmr01_20180510_070000V.stat<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> >
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> >
>
> EnsmStatConfig_APCP24<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> >
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> >
>
> met_stat_analysis_mon_acc_phist.sh<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> >
>
>
>
>
> Hi All,
>
> I want to use Stat_analysis to get seasonal mean rank
histogram, I
> find even if I set ORANK = true in ensemble_stat configure file, the
ORANK
> can't be found in ensemble**.stat file, and it causes the error
information
> when I run stat_analysis mode:
>
> WARNING: do_job_aggr_stat() -> no matching STAT lines found for job:
-job
> aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type
> ORANK -dump_row
> /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
-out_line_type
> PHIST -out_bin_size 0.05000
>
> *****************************
>
> But I have separated files as:
> ensemble_stat_ensmr01_20180510_060000V_orank.nc,
>
> ensemble_stat_ensmr01_20180510_060000V_phist.txt,
>
> ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
>
> Do you know what's issue of my case?
>
> I attached the ensemble_stat configure file, ensemble_stat**.stat
and
> script of aggreate rhist with Stat_analysis.
>
> Thank you very much!
>
> Mandy
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: Zhang, Mandy
Time: Tue Jul 31 16:35:09 2018
Hi Julie
thank you for letting me know!
Mandy
________________________________
From: Julie Prestopnik via RT <met_help at ucar.edu>
Sent: Tuesday, July 31, 2018 4:43:29 PM
To: Zhang, Mandy
Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist with STAT_analysis
Hi Mandy. Our MET expert is out of the office this week.
Unfortunately, I
am unable to help with this issue. I will have our MET expert follow
up
next week.
Julie
On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu>
wrote:
>
> Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> Transaction: Ticket created by Man.Zhang-1 at ou.edu
> Queue: met_help
> Subject: missing Orank when calculate rhist with STAT_analysis
> Owner: Nobody
> Requestors: Man.Zhang-1 at ou.edu
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
>
> Zhang, Mandy has shared OneDrive for Business files with you. To
view
> them, click the links below.
>
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> >
>
> ensemble_stat_ensmr01_20180510_070000V.stat<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> >
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> >
>
> EnsmStatConfig_APCP24<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> >
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> >
>
> met_stat_analysis_mon_acc_phist.sh<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> >
>
>
>
>
> Hi All,
>
> I want to use Stat_analysis to get seasonal mean rank
histogram, I
> find even if I set ORANK = true in ensemble_stat configure file, the
ORANK
> can't be found in ensemble**.stat file, and it causes the error
information
> when I run stat_analysis mode:
>
> WARNING: do_job_aggr_stat() -> no matching STAT lines found for job:
-job
> aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type
> ORANK -dump_row
> /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
-out_line_type
> PHIST -out_bin_size 0.05000
>
> *****************************
>
> But I have separated files as:
> ensemble_stat_ensmr01_20180510_060000V_orank.nc,
>
> ensemble_stat_ensmr01_20180510_060000V_phist.txt,
>
> ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
>
> Do you know what's issue of my case?
>
> I attached the ensemble_stat configure file, ensemble_stat**.stat
and
> script of aggreate rhist with Stat_analysis.
>
> Thank you very much!
>
> Mandy
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: John Halley Gotway
Time: Tue Aug 07 13:04:29 2018
Hi Mandy,
I see you have a question about processing the _orank.nc output file
from
Ensemble-Stat. First, let me briefly review the output lines type
from
Ensemble-Stat...
The ORANK line type stands for observation rank and contains the
observation value and all of the ensemble member values corresponding
to
that observation. Ensemble-Stat can be run to verify against point
observation and/or gridded observations. When verifying with point
observations, the ORANK information is written to an ASCII output
file.
But when verifying with gridded observations, the ORANK information is
written to a NetCDF output file. With gridded obs, there are usually
so
many points that it isn't realistic to write them all to ASCII output.
The RHIST line type is always written the the ASCII output of
Ensemble-Stat
and contains the counts for the binned observation ranks. Basically,
it's
a histogram summarizing the observation ranks across all the verifying
observation points.
Next, STAT-Analysis is designed to only process the ASCII output of
Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't know how
to
process the _orank.nc files.
I see that you'd ultimately like to compute a single ranked histogram
over
a whole month of data. There are 2 ways of doing this:
(1) Aggregate the ORANK lines together into one big RHIST output line
(this is what you're trying but it's not working since you don't have
ORANK
lines in your ASCII output).
(2) Simple aggregate the RHIST lines together over the whole month.
Option 2 is much simpler and requires much less data to process.
Please
try running this job:
-job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type RHIST
And you can try the same thing with the PHIST lines:
-job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type PHIST
Does that do what you want it to do?
Thanks,
John
On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Hi Julie
>
> thank you for letting me know!
>
> Mandy
>
> ________________________________
> From: Julie Prestopnik via RT <met_help at ucar.edu>
> Sent: Tuesday, July 31, 2018 4:43:29 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Hi Mandy. Our MET expert is out of the office this week.
Unfortunately, I
> am unable to help with this issue. I will have our MET expert
follow up
> next week.
>
> Julie
>
> On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > Queue: met_help
> > Subject: missing Orank when calculate rhist with
STAT_analysis
> > Owner: Nobody
> > Requestors: Man.Zhang-1 at ou.edu
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> >
> > Zhang, Mandy has shared OneDrive for Business files with you. To
view
> > them, click the links below.
> >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> >
> > ensemble_stat_ensmr01_20180510_070000V.stat<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> >
> > EnsmStatConfig_APCP24<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> >
> > met_stat_analysis_mon_acc_phist.sh<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> >
> >
> >
> >
> > Hi All,
> >
> > I want to use Stat_analysis to get seasonal mean rank
histogram, I
> > find even if I set ORANK = true in ensemble_stat configure file,
the
> ORANK
> > can't be found in ensemble**.stat file, and it causes the error
> information
> > when I run stat_analysis mode:
> >
> > WARNING: do_job_aggr_stat() -> no matching STAT lines found for
job: -job
> > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> -line_type
> > ORANK -dump_row
> > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> -out_line_type
> > PHIST -out_bin_size 0.05000
> >
> > *****************************
> >
> > But I have separated files as:
> > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> >
> > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> >
> > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> >
> > Do you know what's issue of my case?
> >
> > I attached the ensemble_stat configure file, ensemble_stat**.stat
and
> > script of aggreate rhist with Stat_analysis.
> >
> > Thank you very much!
> >
> > Mandy
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: Zhang, Mandy
Time: Tue Aug 07 13:30:56 2018
Zhang, Mandy has shared OneDrive for Business files with you. To view
them, click the links below.
<https://sooners-my.sharepoint.com/:i:/g/personal/man_zhang-
1_ou_edu/EY_s1DX2BcxEpLijepD0EZkB8icU9uWFORAL-VtAsB3OMA>
[https://r1.res.office365.com/owa/prem/images/dc-
png_20.png]<https://sooners-
my.sharepoint.com/:i:/g/personal/man_zhang-
1_ou_edu/EY_s1DX2BcxEpLijepD0EZkB8icU9uWFORAL-VtAsB3OMA>
2018_sfcr01_ensm_rhist.png<https://sooners-
my.sharepoint.com/:i:/g/personal/man_zhang-
1_ou_edu/EY_s1DX2BcxEpLijepD0EZkB8icU9uWFORAL-VtAsB3OMA>
<https://sooners-my.sharepoint.com/:i:/g/personal/man_zhang-
1_ou_edu/Ed9CDSZ_xalMoX9R8EtVmygBU6e8JeRfltOgPyMZDHUjgQ>
[https://r1.res.office365.com/owa/prem/images/dc-
png_20.png]<https://sooners-
my.sharepoint.com/:i:/g/personal/man_zhang-
1_ou_edu/Ed9CDSZ_xalMoX9R8EtVmygBU6e8JeRfltOgPyMZDHUjgQ>
20180515_sfcr01_ensm_rhist_12hr.png<https://sooners-
my.sharepoint.com/:i:/g/personal/man_zhang-
1_ou_edu/Ed9CDSZ_xalMoX9R8EtVmygBU6e8JeRfltOgPyMZDHUjgQ>
Hi John,
The methods you told me is what I did before, I use command:
-job aggregate_stat -line_type PHIST -out_line_type PHIST -line-
fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I
have -out_line_type PHIST.
But results of whole period looks wried and the single case is
reasonable. please find my attached figs.
And if I want to calculate the CRPS score, does stat_analysis work
with it?
Thank you!
Mandy
________________________________
From: John Halley Gotway via RT <met_help at ucar.edu>
Sent: Tuesday, August 7, 2018 2:04:29 PM
To: Zhang, Mandy
Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist with STAT_analysis
Hi Mandy,
I see you have a question about processing the _orank.nc output file
from
Ensemble-Stat. First, let me briefly review the output lines type
from
Ensemble-Stat...
The ORANK line type stands for observation rank and contains the
observation value and all of the ensemble member values corresponding
to
that observation. Ensemble-Stat can be run to verify against point
observation and/or gridded observations. When verifying with point
observations, the ORANK information is written to an ASCII output
file.
But when verifying with gridded observations, the ORANK information is
written to a NetCDF output file. With gridded obs, there are usually
so
many points that it isn't realistic to write them all to ASCII output.
The RHIST line type is always written the the ASCII output of
Ensemble-Stat
and contains the counts for the binned observation ranks. Basically,
it's
a histogram summarizing the observation ranks across all the verifying
observation points.
Next, STAT-Analysis is designed to only process the ASCII output of
Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't know how
to
process the _orank.nc files.
I see that you'd ultimately like to compute a single ranked histogram
over
a whole month of data. There are 2 ways of doing this:
(1) Aggregate the ORANK lines together into one big RHIST output line
(this is what you're trying but it's not working since you don't have
ORANK
lines in your ASCII output).
(2) Simple aggregate the RHIST lines together over the whole month.
Option 2 is much simpler and requires much less data to process.
Please
try running this job:
-job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type RHIST
And you can try the same thing with the PHIST lines:
-job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type PHIST
Does that do what you want it to do?
Thanks,
John
On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Hi Julie
>
> thank you for letting me know!
>
> Mandy
>
> ________________________________
> From: Julie Prestopnik via RT <met_help at ucar.edu>
> Sent: Tuesday, July 31, 2018 4:43:29 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Hi Mandy. Our MET expert is out of the office this week.
Unfortunately, I
> am unable to help with this issue. I will have our MET expert
follow up
> next week.
>
> Julie
>
> On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > Queue: met_help
> > Subject: missing Orank when calculate rhist with
STAT_analysis
> > Owner: Nobody
> > Requestors: Man.Zhang-1 at ou.edu
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> >
> > Zhang, Mandy has shared OneDrive for Business files with you. To
view
> > them, click the links below.
> >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> >
> > ensemble_stat_ensmr01_20180510_070000V.stat<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> >
> > EnsmStatConfig_APCP24<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> >
> > met_stat_analysis_mon_acc_phist.sh<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> >
> >
> >
> >
> > Hi All,
> >
> > I want to use Stat_analysis to get seasonal mean rank
histogram, I
> > find even if I set ORANK = true in ensemble_stat configure file,
the
> ORANK
> > can't be found in ensemble**.stat file, and it causes the error
> information
> > when I run stat_analysis mode:
> >
> > WARNING: do_job_aggr_stat() -> no matching STAT lines found for
job: -job
> > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> -line_type
> > ORANK -dump_row
> > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> -out_line_type
> > PHIST -out_bin_size 0.05000
> >
> > *****************************
> >
> > But I have separated files as:
> > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> >
> > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> >
> > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> >
> > Do you know what's issue of my case?
> >
> > I attached the ensemble_stat configure file, ensemble_stat**.stat
and
> > script of aggreate rhist with Stat_analysis.
> >
> > Thank you very much!
> >
> > Mandy
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: Zhang, Mandy
Time: Tue Aug 07 13:35:58 2018
Sorry, my job command is -job aggregate -line_type PHIST
-out_line_type PHIST -line-fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST, only difference is I have -out_line_type PHIST.
not aggregate_stat
Mandy
________________________________
From: Zhang, Mandy
Sent: Tuesday, August 7, 2018 2:30:47 PM
To: met_help at ucar.edu
Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist with STAT_analysis
Hi John,
The methods you told me is what I did before, I use command:
-job aggregate_stat -line_type PHIST -out_line_type PHIST -line-
fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I
have -out_line_type PHIST.
But results of whole period looks wried and the single case is
reasonable. please find my attached figs.
And if I want to calculate the CRPS score, does stat_analysis work
with it?
Thank you!
Mandy
________________________________
From: John Halley Gotway via RT <met_help at ucar.edu>
Sent: Tuesday, August 7, 2018 2:04:29 PM
To: Zhang, Mandy
Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist with STAT_analysis
Hi Mandy,
I see you have a question about processing the _orank.nc output file
from
Ensemble-Stat. First, let me briefly review the output lines type
from
Ensemble-Stat...
The ORANK line type stands for observation rank and contains the
observation value and all of the ensemble member values corresponding
to
that observation. Ensemble-Stat can be run to verify against point
observation and/or gridded observations. When verifying with point
observations, the ORANK information is written to an ASCII output
file.
But when verifying with gridded observations, the ORANK information is
written to a NetCDF output file. With gridded obs, there are usually
so
many points that it isn't realistic to write them all to ASCII output.
The RHIST line type is always written the the ASCII output of
Ensemble-Stat
and contains the counts for the binned observation ranks. Basically,
it's
a histogram summarizing the observation ranks across all the verifying
observation points.
Next, STAT-Analysis is designed to only process the ASCII output of
Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't know how
to
process the _orank.nc files.
I see that you'd ultimately like to compute a single ranked histogram
over
a whole month of data. There are 2 ways of doing this:
(1) Aggregate the ORANK lines together into one big RHIST output line
(this is what you're trying but it's not working since you don't have
ORANK
lines in your ASCII output).
(2) Simple aggregate the RHIST lines together over the whole month.
Option 2 is much simpler and requires much less data to process.
Please
try running this job:
-job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type RHIST
And you can try the same thing with the PHIST lines:
-job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
-line_type PHIST
Does that do what you want it to do?
Thanks,
John
On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Hi Julie
>
> thank you for letting me know!
>
> Mandy
>
> ________________________________
> From: Julie Prestopnik via RT <met_help at ucar.edu>
> Sent: Tuesday, July 31, 2018 4:43:29 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Hi Mandy. Our MET expert is out of the office this week.
Unfortunately, I
> am unable to help with this issue. I will have our MET expert
follow up
> next week.
>
> Julie
>
> On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > Queue: met_help
> > Subject: missing Orank when calculate rhist with
STAT_analysis
> > Owner: Nobody
> > Requestors: Man.Zhang-1 at ou.edu
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> >
> > Zhang, Mandy has shared OneDrive for Business files with you. To
view
> > them, click the links below.
> >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> >
> > ensemble_stat_ensmr01_20180510_070000V.stat<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> >
> > EnsmStatConfig_APCP24<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > >
> >
> > <
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> > [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> >
> > met_stat_analysis_mon_acc_phist.sh<
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > >
> >
> >
> >
> >
> > Hi All,
> >
> > I want to use Stat_analysis to get seasonal mean rank
histogram, I
> > find even if I set ORANK = true in ensemble_stat configure file,
the
> ORANK
> > can't be found in ensemble**.stat file, and it causes the error
> information
> > when I run stat_analysis mode:
> >
> > WARNING: do_job_aggr_stat() -> no matching STAT lines found for
job: -job
> > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> -line_type
> > ORANK -dump_row
> > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> -out_line_type
> > PHIST -out_bin_size 0.05000
> >
> > *****************************
> >
> > But I have separated files as:
> > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> >
> > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> >
> > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> >
> > Do you know what's issue of my case?
> >
> > I attached the ensemble_stat configure file, ensemble_stat**.stat
and
> > script of aggreate rhist with Stat_analysis.
> >
> > Thank you very much!
> >
> > Mandy
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: John Halley Gotway
Time: Tue Aug 07 15:24:39 2018
Mandy,
In met-7.0, the CRPS value is included in the RHIST output line type.
I've
attached a sample .stat output file from Ensemble-Stat. This is
created
when you run "make test" for met-7.0.
And here's a job I ran on this file to aggregate multiple RHIST lines
together:
/usr/local/met-7.0/bin/stat_analysis \
-lookin ensemble_stat_20100101_120000V.stat \
-job aggregate -line_type RHIST \
-vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
-obtype ADPSFC
COL_NAME: TOTAL CRPS IGN N_RANK CRPSS SPREAD RANK_1 RANK_2
RANK_3
RANK_4 RANK_5 RANK_6
RANK_7
RHIST: 301 0.30688 0.71961 7 NA 0.072046 39 44
38 35 45 52 48
Notice that the output *DOES* include an aggregated CRPS value.
Next, switch the RHIST to PHIST in the job command:
/usr/local/met-7.0/bin/stat_analysis \
-lookin ensemble_stat_20100101_120000V.stat \
-job aggregate -line_type PHIST \
-vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
-obtype ADPSFC
COL_NAME: TOTAL BIN_SIZE N_BIN BIN_1 BIN_2 BIN_3 BIN_4 BIN_5 BIN_6
BIN_7
BIN_8 BIN_9 BIN_10 BIN_11 BIN_12 BIN_13 BIN_14 BIN_15 BIN_16 BIN_17
BIN_18
BIN_19 BIN_20
PHIST: 36 0.05 20 0 0 0 3 3 7
20 0 0 0 0 0 0 0 0 0 0
0 1 2
Look at the result.
(1) Instead of getting a total of 301 (like in the RHIST), we only get
a
total of 36. That difference is caused by 0 precip. When all the
ensemble members and the observation are 0, the probability integral
transform (PIT) value is undefined.
(2) Notice also that we get a seemingly odd spike at BIN_7... 20 of
the 36
points fall into BIN_7. This happens when the obs value is 0 and the
one
of the ensemble members is non-zero.
When we compute the rank histogram and there is a "tie" in the ranks,
the
code randomly picks a rank from the available ones. There is no
random
picking when computing the PHIST. For truly continuous variables,
like
temperature, the RHIST and PHIST tell you essentially the same thing.
But
for discontinuous variables, like precip, the RHIST and PHIST lines
treat 0
precip differently.
When evaluating precip in Ensemble-Stat, it may be wise to only
include
observation of precip greater than zero. You'd do this in the
Ensemble-Stat config file by setting:
obs_thresh = [ >0 ];
Thanks,
John
On Tue, Aug 7, 2018 at 1:36 PM Zhang, Mandy via RT <met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Sorry, my job command is -job aggregate -line_type PHIST
-out_line_type
> PHIST -line-fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST, only
> difference is I have -out_line_type PHIST.
>
> not aggregate_stat
>
> Mandy
>
> ________________________________
> From: Zhang, Mandy
> Sent: Tuesday, August 7, 2018 2:30:47 PM
> To: met_help at ucar.edu
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
>
> Hi John,
>
> The methods you told me is what I did before, I use command:
>
> -job aggregate_stat -line_type PHIST -out_line_type PHIST -line-
fcst_var
> APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I have
> -out_line_type PHIST.
>
> But results of whole period looks wried and the single case is
> reasonable. please find my attached figs.
>
> And if I want to calculate the CRPS score, does stat_analysis
work
> with it?
>
> Thank you!
>
> Mandy
>
> ________________________________
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Tuesday, August 7, 2018 2:04:29 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Hi Mandy,
>
> I see you have a question about processing the _orank.nc output file
from
> Ensemble-Stat. First, let me briefly review the output lines type
from
> Ensemble-Stat...
>
> The ORANK line type stands for observation rank and contains the
> observation value and all of the ensemble member values
corresponding to
> that observation. Ensemble-Stat can be run to verify against point
> observation and/or gridded observations. When verifying with point
> observations, the ORANK information is written to an ASCII output
file.
> But when verifying with gridded observations, the ORANK information
is
> written to a NetCDF output file. With gridded obs, there are
usually so
> many points that it isn't realistic to write them all to ASCII
output.
>
> The RHIST line type is always written the the ASCII output of
Ensemble-Stat
> and contains the counts for the binned observation ranks.
Basically, it's
> a histogram summarizing the observation ranks across all the
verifying
> observation points.
>
> Next, STAT-Analysis is designed to only process the ASCII output of
> Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't know
how to
> process the _orank.nc files.
>
> I see that you'd ultimately like to compute a single ranked
histogram over
> a whole month of data. There are 2 ways of doing this:
> (1) Aggregate the ORANK lines together into one big RHIST output
line
> (this is what you're trying but it's not working since you don't
have ORANK
> lines in your ASCII output).
> (2) Simple aggregate the RHIST lines together over the whole month.
>
> Option 2 is much simpler and requires much less data to process.
Please
> try running this job:
> -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> -line_type RHIST
>
> And you can try the same thing with the PHIST lines:
> -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> -line_type PHIST
>
> Does that do what you want it to do?
>
> Thanks,
> John
>
> On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> > Hi Julie
> >
> > thank you for letting me know!
> >
> > Mandy
> >
> > ________________________________
> > From: Julie Prestopnik via RT <met_help at ucar.edu>
> > Sent: Tuesday, July 31, 2018 4:43:29 PM
> > To: Zhang, Mandy
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> > Hi Mandy. Our MET expert is out of the office this week.
> Unfortunately, I
> > am unable to help with this issue. I will have our MET expert
follow up
> > next week.
> >
> > Julie
> >
> > On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > > Queue: met_help
> > > Subject: missing Orank when calculate rhist with
STAT_analysis
> > > Owner: Nobody
> > > Requestors: Man.Zhang-1 at ou.edu
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
> >
> > >
> > >
> > > Zhang, Mandy has shared OneDrive for Business files with you. To
view
> > > them, click the links below.
> > >
> > >
> > > <
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > >
> > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > >
> > >
> > > ensemble_stat_ensmr01_20180510_070000V.stat<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > >
> > >
> > > <
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > >
> > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > >
> > >
> > > EnsmStatConfig_APCP24<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > >
> > >
> > > <
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > >
> > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > >
> > >
> > > met_stat_analysis_mon_acc_phist.sh<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > >
> > >
> > >
> > >
> > >
> > > Hi All,
> > >
> > > I want to use Stat_analysis to get seasonal mean rank
histogram, I
> > > find even if I set ORANK = true in ensemble_stat configure file,
the
> > ORANK
> > > can't be found in ensemble**.stat file, and it causes the error
> > information
> > > when I run stat_analysis mode:
> > >
> > > WARNING: do_job_aggr_stat() -> no matching STAT lines found for
job:
> -job
> > > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> > -line_type
> > > ORANK -dump_row
> > > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> > -out_line_type
> > > PHIST -out_bin_size 0.05000
> > >
> > > *****************************
> > >
> > > But I have separated files as:
> > > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> > >
> > > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> > >
> > > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> > >
> > > Do you know what's issue of my case?
> > >
> > > I attached the ensemble_stat configure file,
ensemble_stat**.stat and
> > > script of aggreate rhist with Stat_analysis.
> > >
> > > Thank you very much!
> > >
> > > Mandy
> > >
> > >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: Zhang, Mandy
Time: Tue Aug 07 16:17:55 2018
Hi John,
I'll try both rhist and set flag of obs_thresh in
Ensem_stat.configure, and I'll send you another ticket about the large
domain computing time, it costs too much time to complete a 60-hour
grid_stat run.
Thank you!
Mandy
________________________________
From: John Halley Gotway via RT <met_help at ucar.edu>
Sent: Tuesday, August 7, 2018 4:24:42 PM
To: Zhang, Mandy
Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist with STAT_analysis
Mandy,
In met-7.0, the CRPS value is included in the RHIST output line type.
I've
attached a sample .stat output file from Ensemble-Stat. This is
created
when you run "make test" for met-7.0.
And here's a job I ran on this file to aggregate multiple RHIST lines
together:
/usr/local/met-7.0/bin/stat_analysis \
-lookin ensemble_stat_20100101_120000V.stat \
-job aggregate -line_type RHIST \
-vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
-obtype ADPSFC
COL_NAME: TOTAL CRPS IGN N_RANK CRPSS SPREAD RANK_1 RANK_2
RANK_3
RANK_4 RANK_5 RANK_6
RANK_7
RHIST: 301 0.30688 0.71961 7 NA 0.072046 39 44
38 35 45 52 48
Notice that the output *DOES* include an aggregated CRPS value.
Next, switch the RHIST to PHIST in the job command:
/usr/local/met-7.0/bin/stat_analysis \
-lookin ensemble_stat_20100101_120000V.stat \
-job aggregate -line_type PHIST \
-vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
-obtype ADPSFC
COL_NAME: TOTAL BIN_SIZE N_BIN BIN_1 BIN_2 BIN_3 BIN_4 BIN_5 BIN_6
BIN_7
BIN_8 BIN_9 BIN_10 BIN_11 BIN_12 BIN_13 BIN_14 BIN_15 BIN_16 BIN_17
BIN_18
BIN_19 BIN_20
PHIST: 36 0.05 20 0 0 0 3 3 7
20 0 0 0 0 0 0 0 0 0 0
0 1 2
Look at the result.
(1) Instead of getting a total of 301 (like in the RHIST), we only get
a
total of 36. That difference is caused by 0 precip. When all the
ensemble members and the observation are 0, the probability integral
transform (PIT) value is undefined.
(2) Notice also that we get a seemingly odd spike at BIN_7... 20 of
the 36
points fall into BIN_7. This happens when the obs value is 0 and the
one
of the ensemble members is non-zero.
When we compute the rank histogram and there is a "tie" in the ranks,
the
code randomly picks a rank from the available ones. There is no
random
picking when computing the PHIST. For truly continuous variables,
like
temperature, the RHIST and PHIST tell you essentially the same thing.
But
for discontinuous variables, like precip, the RHIST and PHIST lines
treat 0
precip differently.
When evaluating precip in Ensemble-Stat, it may be wise to only
include
observation of precip greater than zero. You'd do this in the
Ensemble-Stat config file by setting:
obs_thresh = [ >0 ];
Thanks,
John
On Tue, Aug 7, 2018 at 1:36 PM Zhang, Mandy via RT <met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Sorry, my job command is -job aggregate -line_type PHIST
-out_line_type
> PHIST -line-fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST, only
> difference is I have -out_line_type PHIST.
>
> not aggregate_stat
>
> Mandy
>
> ________________________________
> From: Zhang, Mandy
> Sent: Tuesday, August 7, 2018 2:30:47 PM
> To: met_help at ucar.edu
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
>
> Hi John,
>
> The methods you told me is what I did before, I use command:
>
> -job aggregate_stat -line_type PHIST -out_line_type PHIST -line-
fcst_var
> APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I have
> -out_line_type PHIST.
>
> But results of whole period looks wried and the single case is
> reasonable. please find my attached figs.
>
> And if I want to calculate the CRPS score, does stat_analysis
work
> with it?
>
> Thank you!
>
> Mandy
>
> ________________________________
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Tuesday, August 7, 2018 2:04:29 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Hi Mandy,
>
> I see you have a question about processing the _orank.nc output file
from
> Ensemble-Stat. First, let me briefly review the output lines type
from
> Ensemble-Stat...
>
> The ORANK line type stands for observation rank and contains the
> observation value and all of the ensemble member values
corresponding to
> that observation. Ensemble-Stat can be run to verify against point
> observation and/or gridded observations. When verifying with point
> observations, the ORANK information is written to an ASCII output
file.
> But when verifying with gridded observations, the ORANK information
is
> written to a NetCDF output file. With gridded obs, there are
usually so
> many points that it isn't realistic to write them all to ASCII
output.
>
> The RHIST line type is always written the the ASCII output of
Ensemble-Stat
> and contains the counts for the binned observation ranks.
Basically, it's
> a histogram summarizing the observation ranks across all the
verifying
> observation points.
>
> Next, STAT-Analysis is designed to only process the ASCII output of
> Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't know
how to
> process the _orank.nc files.
>
> I see that you'd ultimately like to compute a single ranked
histogram over
> a whole month of data. There are 2 ways of doing this:
> (1) Aggregate the ORANK lines together into one big RHIST output
line
> (this is what you're trying but it's not working since you don't
have ORANK
> lines in your ASCII output).
> (2) Simple aggregate the RHIST lines together over the whole month.
>
> Option 2 is much simpler and requires much less data to process.
Please
> try running this job:
> -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> -line_type RHIST
>
> And you can try the same thing with the PHIST lines:
> -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> -line_type PHIST
>
> Does that do what you want it to do?
>
> Thanks,
> John
>
> On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> > Hi Julie
> >
> > thank you for letting me know!
> >
> > Mandy
> >
> > ________________________________
> > From: Julie Prestopnik via RT <met_help at ucar.edu>
> > Sent: Tuesday, July 31, 2018 4:43:29 PM
> > To: Zhang, Mandy
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> > Hi Mandy. Our MET expert is out of the office this week.
> Unfortunately, I
> > am unable to help with this issue. I will have our MET expert
follow up
> > next week.
> >
> > Julie
> >
> > On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > > Queue: met_help
> > > Subject: missing Orank when calculate rhist with
STAT_analysis
> > > Owner: Nobody
> > > Requestors: Man.Zhang-1 at ou.edu
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
> >
> > >
> > >
> > > Zhang, Mandy has shared OneDrive for Business files with you. To
view
> > > them, click the links below.
> > >
> > >
> > > <
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > >
> > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > >
> > >
> > > ensemble_stat_ensmr01_20180510_070000V.stat<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > >
> > >
> > > <
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > >
> > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > >
> > >
> > > EnsmStatConfig_APCP24<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > >
> > >
> > > <
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > >
> > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > >
> > >
> > > met_stat_analysis_mon_acc_phist.sh<
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > >
> > >
> > >
> > >
> > >
> > > Hi All,
> > >
> > > I want to use Stat_analysis to get seasonal mean rank
histogram, I
> > > find even if I set ORANK = true in ensemble_stat configure file,
the
> > ORANK
> > > can't be found in ensemble**.stat file, and it causes the error
> > information
> > > when I run stat_analysis mode:
> > >
> > > WARNING: do_job_aggr_stat() -> no matching STAT lines found for
job:
> -job
> > > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> > -line_type
> > > ORANK -dump_row
> > > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> > -out_line_type
> > > PHIST -out_bin_size 0.05000
> > >
> > > *****************************
> > >
> > > But I have separated files as:
> > > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> > >
> > > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> > >
> > > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> > >
> > > Do you know what's issue of my case?
> > >
> > > I attached the ensemble_stat configure file,
ensemble_stat**.stat and
> > > script of aggreate rhist with Stat_analysis.
> > >
> > > Thank you very much!
> > >
> > > Mandy
> > >
> > >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: John Halley Gotway
Time: Tue Aug 07 16:24:07 2018
Mandy,
OK, please do see the bugfix posted for met-7.0 on 7/24/18:
https://dtcenter.org/met/users/support/known_issues/METv7.0/index.php
Perhaps that's the issue causing the slow runtime?
John
On Tue, Aug 7, 2018 at 4:18 PM Zhang, Mandy via RT <met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Hi John,
>
> I'll try both rhist and set flag of obs_thresh in
> Ensem_stat.configure, and I'll send you another ticket about the
large
> domain computing time, it costs too much time to complete a 60-hour
> grid_stat run.
>
> Thank you!
>
> Mandy
>
> ________________________________
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Tuesday, August 7, 2018 4:24:42 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Mandy,
>
> In met-7.0, the CRPS value is included in the RHIST output line
type. I've
> attached a sample .stat output file from Ensemble-Stat. This is
created
> when you run "make test" for met-7.0.
>
> And here's a job I ran on this file to aggregate multiple RHIST
lines
> together:
>
> /usr/local/met-7.0/bin/stat_analysis \
> -lookin ensemble_stat_20100101_120000V.stat \
> -job aggregate -line_type RHIST \
> -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> -obtype ADPSFC
>
> COL_NAME: TOTAL CRPS IGN N_RANK CRPSS SPREAD RANK_1 RANK_2
RANK_3
> RANK_4 RANK_5 RANK_6
> RANK_7
>
> RHIST: 301 0.30688 0.71961 7 NA 0.072046 39 44
> 38 35 45 52 48
>
> Notice that the output *DOES* include an aggregated CRPS value.
>
> Next, switch the RHIST to PHIST in the job command:
>
> /usr/local/met-7.0/bin/stat_analysis \
> -lookin ensemble_stat_20100101_120000V.stat \
> -job aggregate -line_type PHIST \
> -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> -obtype ADPSFC
>
> COL_NAME: TOTAL BIN_SIZE N_BIN BIN_1 BIN_2 BIN_3 BIN_4 BIN_5 BIN_6
BIN_7
> BIN_8 BIN_9 BIN_10 BIN_11 BIN_12 BIN_13 BIN_14 BIN_15 BIN_16 BIN_17
BIN_18
> BIN_19 BIN_20
> PHIST: 36 0.05 20 0 0 0 3 3 7
> 20 0 0 0 0 0 0 0 0 0
0
> 0 1 2
>
> Look at the result.
>
> (1) Instead of getting a total of 301 (like in the RHIST), we only
get a
> total of 36. That difference is caused by 0 precip. When all the
> ensemble members and the observation are 0, the probability integral
> transform (PIT) value is undefined.
> (2) Notice also that we get a seemingly odd spike at BIN_7... 20 of
the 36
> points fall into BIN_7. This happens when the obs value is 0 and
the one
> of the ensemble members is non-zero.
>
> When we compute the rank histogram and there is a "tie" in the
ranks, the
> code randomly picks a rank from the available ones. There is no
random
> picking when computing the PHIST. For truly continuous variables,
like
> temperature, the RHIST and PHIST tell you essentially the same
thing. But
> for discontinuous variables, like precip, the RHIST and PHIST lines
treat 0
> precip differently.
>
> When evaluating precip in Ensemble-Stat, it may be wise to only
include
> observation of precip greater than zero. You'd do this in the
> Ensemble-Stat config file by setting:
> obs_thresh = [ >0 ];
>
> Thanks,
> John
>
>
>
> On Tue, Aug 7, 2018 at 1:36 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> > Sorry, my job command is -job aggregate -line_type PHIST
-out_line_type
> > PHIST -line-fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST, only
> > difference is I have -out_line_type PHIST.
> >
> > not aggregate_stat
> >
> > Mandy
> >
> > ________________________________
> > From: Zhang, Mandy
> > Sent: Tuesday, August 7, 2018 2:30:47 PM
> > To: met_help at ucar.edu
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> >
> > Hi John,
> >
> > The methods you told me is what I did before, I use command:
> >
> > -job aggregate_stat -line_type PHIST -out_line_type PHIST -line-
fcst_var
> > APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I have
> > -out_line_type PHIST.
> >
> > But results of whole period looks wried and the single case is
> > reasonable. please find my attached figs.
> >
> > And if I want to calculate the CRPS score, does stat_analysis
work
> > with it?
> >
> > Thank you!
> >
> > Mandy
> >
> > ________________________________
> > From: John Halley Gotway via RT <met_help at ucar.edu>
> > Sent: Tuesday, August 7, 2018 2:04:29 PM
> > To: Zhang, Mandy
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> > Hi Mandy,
> >
> > I see you have a question about processing the _orank.nc output
file
> from
> > Ensemble-Stat. First, let me briefly review the output lines type
from
> > Ensemble-Stat...
> >
> > The ORANK line type stands for observation rank and contains the
> > observation value and all of the ensemble member values
corresponding to
> > that observation. Ensemble-Stat can be run to verify against
point
> > observation and/or gridded observations. When verifying with
point
> > observations, the ORANK information is written to an ASCII output
file.
> > But when verifying with gridded observations, the ORANK
information is
> > written to a NetCDF output file. With gridded obs, there are
usually so
> > many points that it isn't realistic to write them all to ASCII
output.
> >
> > The RHIST line type is always written the the ASCII output of
> Ensemble-Stat
> > and contains the counts for the binned observation ranks.
Basically,
> it's
> > a histogram summarizing the observation ranks across all the
verifying
> > observation points.
> >
> > Next, STAT-Analysis is designed to only process the ASCII output
of
> > Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't know
how to
> > process the _orank.nc files.
> >
> > I see that you'd ultimately like to compute a single ranked
histogram
> over
> > a whole month of data. There are 2 ways of doing this:
> > (1) Aggregate the ORANK lines together into one big RHIST output
line
> > (this is what you're trying but it's not working since you don't
have
> ORANK
> > lines in your ASCII output).
> > (2) Simple aggregate the RHIST lines together over the whole
month.
> >
> > Option 2 is much simpler and requires much less data to process.
Please
> > try running this job:
> > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> > -line_type RHIST
> >
> > And you can try the same thing with the PHIST lines:
> > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> > -line_type PHIST
> >
> > Does that do what you want it to do?
> >
> > Thanks,
> > John
> >
> > On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> > >
> > > Hi Julie
> > >
> > > thank you for letting me know!
> > >
> > > Mandy
> > >
> > > ________________________________
> > > From: Julie Prestopnik via RT <met_help at ucar.edu>
> > > Sent: Tuesday, July 31, 2018 4:43:29 PM
> > > To: Zhang, Mandy
> > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> rhist
> > > with STAT_analysis
> > >
> > > Hi Mandy. Our MET expert is out of the office this week.
> > Unfortunately, I
> > > am unable to help with this issue. I will have our MET expert
follow
> up
> > > next week.
> > >
> > > Julie
> > >
> > > On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu
> >
> > > wrote:
> > >
> > > >
> > > > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > > > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > > > Queue: met_help
> > > > Subject: missing Orank when calculate rhist with
STAT_analysis
> > > > Owner: Nobody
> > > > Requestors: Man.Zhang-1 at ou.edu
> > > > Status: new
> > > > Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
> > >
> > > >
> > > >
> > > > Zhang, Mandy has shared OneDrive for Business files with you.
To view
> > > > them, click the links below.
> > > >
> > > >
> > > > <
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > >
> > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > >
> > > >
> > > > ensemble_stat_ensmr01_20180510_070000V.stat<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > >
> > > >
> > > > <
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > >
> > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > >
> > > >
> > > > EnsmStatConfig_APCP24<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > >
> > > >
> > > > <
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > >
> > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > >
> > > >
> > > > met_stat_analysis_mon_acc_phist.sh<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Hi All,
> > > >
> > > > I want to use Stat_analysis to get seasonal mean rank
> histogram, I
> > > > find even if I set ORANK = true in ensemble_stat configure
file, the
> > > ORANK
> > > > can't be found in ensemble**.stat file, and it causes the
error
> > > information
> > > > when I run stat_analysis mode:
> > > >
> > > > WARNING: do_job_aggr_stat() -> no matching STAT lines found
for job:
> > -job
> > > > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > -line_type
> > > > ORANK -dump_row
> > > > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> > > -out_line_type
> > > > PHIST -out_bin_size 0.05000
> > > >
> > > > *****************************
> > > >
> > > > But I have separated files as:
> > > > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> > > >
> > > > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> > > >
> > > > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> > > >
> > > > Do you know what's issue of my case?
> > > >
> > > > I attached the ensemble_stat configure file,
ensemble_stat**.stat and
> > > > script of aggreate rhist with Stat_analysis.
> > > >
> > > > Thank you very much!
> > > >
> > > > Mandy
> > > >
> > > >
> > >
> > >
> > >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: Zhang, Mandy
Time: Wed Aug 08 11:04:06 2018
Zhang, Mandy has shared a OneDrive for Business file with you. To view
it, click the link below.
<https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg>
[https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<https://sooners-
my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg>
make-met7.0-bug<https://sooners-
my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg>
Hi John,
I run point_stat with latest version MET-7.0-bugfix , but it
looks a limited improvement of runtime, a single run still needs 10
minutes to complete.
And when I compile met-7.0 with PGI compiler, there's a error in
/home/mzhang/tool/met-7.0/src/tools/other/pb2nc/pb2nc.cc,
****************
"pb2nc.cc", line 576: error: no instance of overloaded function
"getline"
matches the argument list
argument types are: (char **, size_t *, FILE *)
while ((read = getline(&line, &len, fp)) != -1) {
******
please find details in the attachment .
Thank you very much!
Mandy
________________________________
From: John Halley Gotway via RT <met_help at ucar.edu>
Sent: Tuesday, August 7, 2018 5:24:07 PM
To: Zhang, Mandy
Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist with STAT_analysis
Mandy,
OK, please do see the bugfix posted for met-7.0 on 7/24/18:
https://dtcenter.org/met/users/support/known_issues/METv7.0/index.php
Perhaps that's the issue causing the slow runtime?
John
On Tue, Aug 7, 2018 at 4:18 PM Zhang, Mandy via RT <met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Hi John,
>
> I'll try both rhist and set flag of obs_thresh in
> Ensem_stat.configure, and I'll send you another ticket about the
large
> domain computing time, it costs too much time to complete a 60-hour
> grid_stat run.
>
> Thank you!
>
> Mandy
>
> ________________________________
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Tuesday, August 7, 2018 4:24:42 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Mandy,
>
> In met-7.0, the CRPS value is included in the RHIST output line
type. I've
> attached a sample .stat output file from Ensemble-Stat. This is
created
> when you run "make test" for met-7.0.
>
> And here's a job I ran on this file to aggregate multiple RHIST
lines
> together:
>
> /usr/local/met-7.0/bin/stat_analysis \
> -lookin ensemble_stat_20100101_120000V.stat \
> -job aggregate -line_type RHIST \
> -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> -obtype ADPSFC
>
> COL_NAME: TOTAL CRPS IGN N_RANK CRPSS SPREAD RANK_1 RANK_2
RANK_3
> RANK_4 RANK_5 RANK_6
> RANK_7
>
> RHIST: 301 0.30688 0.71961 7 NA 0.072046 39 44
> 38 35 45 52 48
>
> Notice that the output *DOES* include an aggregated CRPS value.
>
> Next, switch the RHIST to PHIST in the job command:
>
> /usr/local/met-7.0/bin/stat_analysis \
> -lookin ensemble_stat_20100101_120000V.stat \
> -job aggregate -line_type PHIST \
> -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> -obtype ADPSFC
>
> COL_NAME: TOTAL BIN_SIZE N_BIN BIN_1 BIN_2 BIN_3 BIN_4 BIN_5 BIN_6
BIN_7
> BIN_8 BIN_9 BIN_10 BIN_11 BIN_12 BIN_13 BIN_14 BIN_15 BIN_16 BIN_17
BIN_18
> BIN_19 BIN_20
> PHIST: 36 0.05 20 0 0 0 3 3 7
> 20 0 0 0 0 0 0 0 0 0
0
> 0 1 2
>
> Look at the result.
>
> (1) Instead of getting a total of 301 (like in the RHIST), we only
get a
> total of 36. That difference is caused by 0 precip. When all the
> ensemble members and the observation are 0, the probability integral
> transform (PIT) value is undefined.
> (2) Notice also that we get a seemingly odd spike at BIN_7... 20 of
the 36
> points fall into BIN_7. This happens when the obs value is 0 and
the one
> of the ensemble members is non-zero.
>
> When we compute the rank histogram and there is a "tie" in the
ranks, the
> code randomly picks a rank from the available ones. There is no
random
> picking when computing the PHIST. For truly continuous variables,
like
> temperature, the RHIST and PHIST tell you essentially the same
thing. But
> for discontinuous variables, like precip, the RHIST and PHIST lines
treat 0
> precip differently.
>
> When evaluating precip in Ensemble-Stat, it may be wise to only
include
> observation of precip greater than zero. You'd do this in the
> Ensemble-Stat config file by setting:
> obs_thresh = [ >0 ];
>
> Thanks,
> John
>
>
>
> On Tue, Aug 7, 2018 at 1:36 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> > Sorry, my job command is -job aggregate -line_type PHIST
-out_line_type
> > PHIST -line-fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST, only
> > difference is I have -out_line_type PHIST.
> >
> > not aggregate_stat
> >
> > Mandy
> >
> > ________________________________
> > From: Zhang, Mandy
> > Sent: Tuesday, August 7, 2018 2:30:47 PM
> > To: met_help at ucar.edu
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> >
> > Hi John,
> >
> > The methods you told me is what I did before, I use command:
> >
> > -job aggregate_stat -line_type PHIST -out_line_type PHIST -line-
fcst_var
> > APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I have
> > -out_line_type PHIST.
> >
> > But results of whole period looks wried and the single case is
> > reasonable. please find my attached figs.
> >
> > And if I want to calculate the CRPS score, does stat_analysis
work
> > with it?
> >
> > Thank you!
> >
> > Mandy
> >
> > ________________________________
> > From: John Halley Gotway via RT <met_help at ucar.edu>
> > Sent: Tuesday, August 7, 2018 2:04:29 PM
> > To: Zhang, Mandy
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> > Hi Mandy,
> >
> > I see you have a question about processing the _orank.nc output
file
> from
> > Ensemble-Stat. First, let me briefly review the output lines type
from
> > Ensemble-Stat...
> >
> > The ORANK line type stands for observation rank and contains the
> > observation value and all of the ensemble member values
corresponding to
> > that observation. Ensemble-Stat can be run to verify against
point
> > observation and/or gridded observations. When verifying with
point
> > observations, the ORANK information is written to an ASCII output
file.
> > But when verifying with gridded observations, the ORANK
information is
> > written to a NetCDF output file. With gridded obs, there are
usually so
> > many points that it isn't realistic to write them all to ASCII
output.
> >
> > The RHIST line type is always written the the ASCII output of
> Ensemble-Stat
> > and contains the counts for the binned observation ranks.
Basically,
> it's
> > a histogram summarizing the observation ranks across all the
verifying
> > observation points.
> >
> > Next, STAT-Analysis is designed to only process the ASCII output
of
> > Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't know
how to
> > process the _orank.nc files.
> >
> > I see that you'd ultimately like to compute a single ranked
histogram
> over
> > a whole month of data. There are 2 ways of doing this:
> > (1) Aggregate the ORANK lines together into one big RHIST output
line
> > (this is what you're trying but it's not working since you don't
have
> ORANK
> > lines in your ASCII output).
> > (2) Simple aggregate the RHIST lines together over the whole
month.
> >
> > Option 2 is much simpler and requires much less data to process.
Please
> > try running this job:
> > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> > -line_type RHIST
> >
> > And you can try the same thing with the PHIST lines:
> > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST
> > -line_type PHIST
> >
> > Does that do what you want it to do?
> >
> > Thanks,
> > John
> >
> > On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> > >
> > > Hi Julie
> > >
> > > thank you for letting me know!
> > >
> > > Mandy
> > >
> > > ________________________________
> > > From: Julie Prestopnik via RT <met_help at ucar.edu>
> > > Sent: Tuesday, July 31, 2018 4:43:29 PM
> > > To: Zhang, Mandy
> > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> rhist
> > > with STAT_analysis
> > >
> > > Hi Mandy. Our MET expert is out of the office this week.
> > Unfortunately, I
> > > am unable to help with this issue. I will have our MET expert
follow
> up
> > > next week.
> > >
> > > Julie
> > >
> > > On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT
<met_help at ucar.edu
> >
> > > wrote:
> > >
> > > >
> > > > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > > > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > > > Queue: met_help
> > > > Subject: missing Orank when calculate rhist with
STAT_analysis
> > > > Owner: Nobody
> > > > Requestors: Man.Zhang-1 at ou.edu
> > > > Status: new
> > > > Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
> > >
> > > >
> > > >
> > > > Zhang, Mandy has shared OneDrive for Business files with you.
To view
> > > > them, click the links below.
> > > >
> > > >
> > > > <
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > >
> > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > >
> > > >
> > > > ensemble_stat_ensmr01_20180510_070000V.stat<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > >
> > > >
> > > > <
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > >
> > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > >
> > > >
> > > > EnsmStatConfig_APCP24<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > >
> > > >
> > > > <
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > >
> > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > >
> > > >
> > > > met_stat_analysis_mon_acc_phist.sh<
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Hi All,
> > > >
> > > > I want to use Stat_analysis to get seasonal mean rank
> histogram, I
> > > > find even if I set ORANK = true in ensemble_stat configure
file, the
> > > ORANK
> > > > can't be found in ensemble**.stat file, and it causes the
error
> > > information
> > > > when I run stat_analysis mode:
> > > >
> > > > WARNING: do_job_aggr_stat() -> no matching STAT lines found
for job:
> > -job
> > > > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > -line_type
> > > > ORANK -dump_row
> > > > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> > > -out_line_type
> > > > PHIST -out_bin_size 0.05000
> > > >
> > > > *****************************
> > > >
> > > > But I have separated files as:
> > > > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> > > >
> > > > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> > > >
> > > > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> > > >
> > > > Do you know what's issue of my case?
> > > >
> > > > I attached the ensemble_stat configure file,
ensemble_stat**.stat and
> > > > script of aggreate rhist with Stat_analysis.
> > > >
> > > > Thank you very much!
> > > >
> > > > Mandy
> > > >
> > > >
> > >
> > >
> > >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: John Halley Gotway
Time: Wed Aug 08 13:19:52 2018
Mandy,
Thanks for letting us know about the PGI compilation issue.
Throughout the
DTC, we've been having increasing problems with the PGI compiler. In
fact,
I believe that both WRF and GSI have stopped supporting it.
We routinely compile with GNU and Intel, so if those are available to
you,
I'd recommend them.
Thanks,
John
On Wed, Aug 8, 2018 at 11:04 AM Zhang, Mandy via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Zhang, Mandy has shared a OneDrive for Business file with you. To
view it,
> click the link below.
>
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg
> >
>
> make-met7.0-bug<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg
> >
>
>
>
>
> Hi John,
>
> I run point_stat with latest version MET-7.0-bugfix , but it
looks a
> limited improvement of runtime, a single run still needs 10 minutes
to
> complete.
>
> And when I compile met-7.0 with PGI compiler, there's a error
in
> /home/mzhang/tool/met-7.0/src/tools/other/pb2nc/pb2nc.cc,
>
> ****************
>
> "pb2nc.cc", line 576: error: no instance of overloaded function
"getline"
> matches the argument list
> argument types are: (char **, size_t *, FILE *)
> while ((read = getline(&line, &len, fp)) != -1) {
> ******
> please find details in the attachment .
> Thank you very much!
> Mandy
>
> ________________________________
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Tuesday, August 7, 2018 5:24:07 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Mandy,
>
> OK, please do see the bugfix posted for met-7.0 on 7/24/18:
>
https://dtcenter.org/met/users/support/known_issues/METv7.0/index.php
>
> Perhaps that's the issue causing the slow runtime?
>
> John
>
> On Tue, Aug 7, 2018 at 4:18 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> > Hi John,
> >
> > I'll try both rhist and set flag of obs_thresh in
> > Ensem_stat.configure, and I'll send you another ticket about the
large
> > domain computing time, it costs too much time to complete a 60-
hour
> > grid_stat run.
> >
> > Thank you!
> >
> > Mandy
> >
> > ________________________________
> > From: John Halley Gotway via RT <met_help at ucar.edu>
> > Sent: Tuesday, August 7, 2018 4:24:42 PM
> > To: Zhang, Mandy
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> > Mandy,
> >
> > In met-7.0, the CRPS value is included in the RHIST output line
type.
> I've
> > attached a sample .stat output file from Ensemble-Stat. This is
created
> > when you run "make test" for met-7.0.
> >
> > And here's a job I ran on this file to aggregate multiple RHIST
lines
> > together:
> >
> > /usr/local/met-7.0/bin/stat_analysis \
> > -lookin ensemble_stat_20100101_120000V.stat \
> > -job aggregate -line_type RHIST \
> > -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> > -obtype ADPSFC
> >
> > COL_NAME: TOTAL CRPS IGN N_RANK CRPSS SPREAD RANK_1
RANK_2
> RANK_3
> > RANK_4 RANK_5 RANK_6
> > RANK_7
> >
> > RHIST: 301 0.30688 0.71961 7 NA 0.072046 39
44
> > 38 35 45 52 48
> >
> > Notice that the output *DOES* include an aggregated CRPS value.
> >
> > Next, switch the RHIST to PHIST in the job command:
> >
> > /usr/local/met-7.0/bin/stat_analysis \
> > -lookin ensemble_stat_20100101_120000V.stat \
> > -job aggregate -line_type PHIST \
> > -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> > -obtype ADPSFC
> >
> > COL_NAME: TOTAL BIN_SIZE N_BIN BIN_1 BIN_2 BIN_3 BIN_4 BIN_5 BIN_6
BIN_7
> > BIN_8 BIN_9 BIN_10 BIN_11 BIN_12 BIN_13 BIN_14 BIN_15 BIN_16
BIN_17
> BIN_18
> > BIN_19 BIN_20
> > PHIST: 36 0.05 20 0 0 0 3 3 7
> > 20 0 0 0 0 0 0 0 0 0
0
> > 0 1 2
> >
> > Look at the result.
> >
> > (1) Instead of getting a total of 301 (like in the RHIST), we only
get a
> > total of 36. That difference is caused by 0 precip. When all
the
> > ensemble members and the observation are 0, the probability
integral
> > transform (PIT) value is undefined.
> > (2) Notice also that we get a seemingly odd spike at BIN_7... 20
of the
> 36
> > points fall into BIN_7. This happens when the obs value is 0 and
the one
> > of the ensemble members is non-zero.
> >
> > When we compute the rank histogram and there is a "tie" in the
ranks, the
> > code randomly picks a rank from the available ones. There is no
random
> > picking when computing the PHIST. For truly continuous variables,
like
> > temperature, the RHIST and PHIST tell you essentially the same
thing.
> But
> > for discontinuous variables, like precip, the RHIST and PHIST
lines
> treat 0
> > precip differently.
> >
> > When evaluating precip in Ensemble-Stat, it may be wise to only
include
> > observation of precip greater than zero. You'd do this in the
> > Ensemble-Stat config file by setting:
> > obs_thresh = [ >0 ];
> >
> > Thanks,
> > John
> >
> >
> >
> > On Tue, Aug 7, 2018 at 1:36 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> > >
> > > Sorry, my job command is -job aggregate -line_type PHIST
-out_line_type
> > > PHIST -line-fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST,
only
> > > difference is I have -out_line_type PHIST.
> > >
> > > not aggregate_stat
> > >
> > > Mandy
> > >
> > > ________________________________
> > > From: Zhang, Mandy
> > > Sent: Tuesday, August 7, 2018 2:30:47 PM
> > > To: met_help at ucar.edu
> > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> rhist
> > > with STAT_analysis
> > >
> > >
> > > Hi John,
> > >
> > > The methods you told me is what I did before, I use
command:
> > >
> > > -job aggregate_stat -line_type PHIST -out_line_type PHIST
> -line-fcst_var
> > > APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I
have
> > > -out_line_type PHIST.
> > >
> > > But results of whole period looks wried and the single case
is
> > > reasonable. please find my attached figs.
> > >
> > > And if I want to calculate the CRPS score, does
stat_analysis work
> > > with it?
> > >
> > > Thank you!
> > >
> > > Mandy
> > >
> > > ________________________________
> > > From: John Halley Gotway via RT <met_help at ucar.edu>
> > > Sent: Tuesday, August 7, 2018 2:04:29 PM
> > > To: Zhang, Mandy
> > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> rhist
> > > with STAT_analysis
> > >
> > > Hi Mandy,
> > >
> > > I see you have a question about processing the _orank.nc output
file
> > from
> > > Ensemble-Stat. First, let me briefly review the output lines
type from
> > > Ensemble-Stat...
> > >
> > > The ORANK line type stands for observation rank and contains the
> > > observation value and all of the ensemble member values
corresponding
> to
> > > that observation. Ensemble-Stat can be run to verify against
point
> > > observation and/or gridded observations. When verifying with
point
> > > observations, the ORANK information is written to an ASCII
output file.
> > > But when verifying with gridded observations, the ORANK
information is
> > > written to a NetCDF output file. With gridded obs, there are
usually
> so
> > > many points that it isn't realistic to write them all to ASCII
output.
> > >
> > > The RHIST line type is always written the the ASCII output of
> > Ensemble-Stat
> > > and contains the counts for the binned observation ranks.
Basically,
> > it's
> > > a histogram summarizing the observation ranks across all the
verifying
> > > observation points.
> > >
> > > Next, STAT-Analysis is designed to only process the ASCII output
of
> > > Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't
know how
> to
> > > process the _orank.nc files.
> > >
> > > I see that you'd ultimately like to compute a single ranked
histogram
> > over
> > > a whole month of data. There are 2 ways of doing this:
> > > (1) Aggregate the ORANK lines together into one big RHIST
output line
> > > (this is what you're trying but it's not working since you don't
have
> > ORANK
> > > lines in your ASCII output).
> > > (2) Simple aggregate the RHIST lines together over the whole
month.
> > >
> > > Option 2 is much simpler and requires much less data to process.
> Please
> > > try running this job:
> > > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > -line_type RHIST
> > >
> > > And you can try the same thing with the PHIST lines:
> > > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > -line_type PHIST
> > >
> > > Does that do what you want it to do?
> > >
> > > Thanks,
> > > John
> > >
> > > On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu
> >
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
>
> > > >
> > > > Hi Julie
> > > >
> > > > thank you for letting me know!
> > > >
> > > > Mandy
> > > >
> > > > ________________________________
> > > > From: Julie Prestopnik via RT <met_help at ucar.edu>
> > > > Sent: Tuesday, July 31, 2018 4:43:29 PM
> > > > To: Zhang, Mandy
> > > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> > rhist
> > > > with STAT_analysis
> > > >
> > > > Hi Mandy. Our MET expert is out of the office this week.
> > > Unfortunately, I
> > > > am unable to help with this issue. I will have our MET expert
follow
> > up
> > > > next week.
> > > >
> > > > Julie
> > > >
> > > > On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT <
> met_help at ucar.edu
> > >
> > > > wrote:
> > > >
> > > > >
> > > > > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > > > > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > > > > Queue: met_help
> > > > > Subject: missing Orank when calculate rhist with
> STAT_analysis
> > > > > Owner: Nobody
> > > > > Requestors: Man.Zhang-1 at ou.edu
> > > > > Status: new
> > > > > Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
> > > >
> > > > >
> > > > >
> > > > > Zhang, Mandy has shared OneDrive for Business files with
you. To
> view
> > > > > them, click the links below.
> > > > >
> > > > >
> > > > > <
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > > >
> > > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > > >
> > > > >
> > > > > ensemble_stat_ensmr01_20180510_070000V.stat<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > > >
> > > > >
> > > > > <
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > > >
> > > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > > >
> > > > >
> > > > > EnsmStatConfig_APCP24<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > > >
> > > > >
> > > > > <
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > > >
> > > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > > >
> > > > >
> > > > > met_stat_analysis_mon_acc_phist.sh<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Hi All,
> > > > >
> > > > > I want to use Stat_analysis to get seasonal mean rank
> > histogram, I
> > > > > find even if I set ORANK = true in ensemble_stat configure
file,
> the
> > > > ORANK
> > > > > can't be found in ensemble**.stat file, and it causes the
error
> > > > information
> > > > > when I run stat_analysis mode:
> > > > >
> > > > > WARNING: do_job_aggr_stat() -> no matching STAT lines found
for
> job:
> > > -job
> > > > > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > > -line_type
> > > > > ORANK -dump_row
> > > > > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> > > > -out_line_type
> > > > > PHIST -out_bin_size 0.05000
> > > > >
> > > > > *****************************
> > > > >
> > > > > But I have separated files as:
> > > > > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> > > > >
> > > > > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> > > > >
> > > > > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> > > > >
> > > > > Do you know what's issue of my case?
> > > > >
> > > > > I attached the ensemble_stat configure file,
ensemble_stat**.stat
> and
> > > > > script of aggreate rhist with Stat_analysis.
> > > > >
> > > > > Thank you very much!
> > > > >
> > > > > Mandy
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> >
> >
> >
>
>
>
------------------------------------------------
Subject: missing Orank when calculate rhist with STAT_analysis
From: Zhang, Mandy
Time: Wed Aug 08 13:31:56 2018
Hi John,
I have met-7.0 with intel compiler, and it works well.
Thank you~
Mandy
________________________________
From: John Halley Gotway via RT <met_help at ucar.edu>
Sent: Wednesday, August 8, 2018 2:19:53 PM
To: Zhang, Mandy
Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist with STAT_analysis
Mandy,
Thanks for letting us know about the PGI compilation issue.
Throughout the
DTC, we've been having increasing problems with the PGI compiler. In
fact,
I believe that both WRF and GSI have stopped supporting it.
We routinely compile with GNU and Intel, so if those are available to
you,
I'd recommend them.
Thanks,
John
On Wed, Aug 8, 2018 at 11:04 AM Zhang, Mandy via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
>
> Zhang, Mandy has shared a OneDrive for Business file with you. To
view it,
> click the link below.
>
>
> <
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg
> >
> [https://r1.res.office365.com/owa/prem/images/dc-generic_20.png]<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg
> >
>
> make-met7.0-bug<
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EQn8AK_5GjJBsEx2lE1y6a0BSBkSFSq7mS3Z9ij_ZU00Xg
> >
>
>
>
>
> Hi John,
>
> I run point_stat with latest version MET-7.0-bugfix , but it
looks a
> limited improvement of runtime, a single run still needs 10 minutes
to
> complete.
>
> And when I compile met-7.0 with PGI compiler, there's a error
in
> /home/mzhang/tool/met-7.0/src/tools/other/pb2nc/pb2nc.cc,
>
> ****************
>
> "pb2nc.cc", line 576: error: no instance of overloaded function
"getline"
> matches the argument list
> argument types are: (char **, size_t *, FILE *)
> while ((read = getline(&line, &len, fp)) != -1) {
> ******
> please find details in the attachment .
> Thank you very much!
> Mandy
>
> ________________________________
> From: John Halley Gotway via RT <met_help at ucar.edu>
> Sent: Tuesday, August 7, 2018 5:24:07 PM
> To: Zhang, Mandy
> Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> with STAT_analysis
>
> Mandy,
>
> OK, please do see the bugfix posted for met-7.0 on 7/24/18:
>
https://dtcenter.org/met/users/support/known_issues/METv7.0/index.php
>
> Perhaps that's the issue causing the slow runtime?
>
> John
>
> On Tue, Aug 7, 2018 at 4:18 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> >
> > Hi John,
> >
> > I'll try both rhist and set flag of obs_thresh in
> > Ensem_stat.configure, and I'll send you another ticket about the
large
> > domain computing time, it costs too much time to complete a 60-
hour
> > grid_stat run.
> >
> > Thank you!
> >
> > Mandy
> >
> > ________________________________
> > From: John Halley Gotway via RT <met_help at ucar.edu>
> > Sent: Tuesday, August 7, 2018 4:24:42 PM
> > To: Zhang, Mandy
> > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when calculate
rhist
> > with STAT_analysis
> >
> > Mandy,
> >
> > In met-7.0, the CRPS value is included in the RHIST output line
type.
> I've
> > attached a sample .stat output file from Ensemble-Stat. This is
created
> > when you run "make test" for met-7.0.
> >
> > And here's a job I ran on this file to aggregate multiple RHIST
lines
> > together:
> >
> > /usr/local/met-7.0/bin/stat_analysis \
> > -lookin ensemble_stat_20100101_120000V.stat \
> > -job aggregate -line_type RHIST \
> > -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> > -obtype ADPSFC
> >
> > COL_NAME: TOTAL CRPS IGN N_RANK CRPSS SPREAD RANK_1
RANK_2
> RANK_3
> > RANK_4 RANK_5 RANK_6
> > RANK_7
> >
> > RHIST: 301 0.30688 0.71961 7 NA 0.072046 39
44
> > 38 35 45 52 48
> >
> > Notice that the output *DOES* include an aggregated CRPS value.
> >
> > Next, switch the RHIST to PHIST in the job command:
> >
> > /usr/local/met-7.0/bin/stat_analysis \
> > -lookin ensemble_stat_20100101_120000V.stat \
> > -job aggregate -line_type PHIST \
> > -vx_mask HUC4_1605,HUC4_1803,HUC4_1804,HUC4_1805,HUC4_1806 \
> > -obtype ADPSFC
> >
> > COL_NAME: TOTAL BIN_SIZE N_BIN BIN_1 BIN_2 BIN_3 BIN_4 BIN_5 BIN_6
BIN_7
> > BIN_8 BIN_9 BIN_10 BIN_11 BIN_12 BIN_13 BIN_14 BIN_15 BIN_16
BIN_17
> BIN_18
> > BIN_19 BIN_20
> > PHIST: 36 0.05 20 0 0 0 3 3 7
> > 20 0 0 0 0 0 0 0 0 0
0
> > 0 1 2
> >
> > Look at the result.
> >
> > (1) Instead of getting a total of 301 (like in the RHIST), we only
get a
> > total of 36. That difference is caused by 0 precip. When all
the
> > ensemble members and the observation are 0, the probability
integral
> > transform (PIT) value is undefined.
> > (2) Notice also that we get a seemingly odd spike at BIN_7... 20
of the
> 36
> > points fall into BIN_7. This happens when the obs value is 0 and
the one
> > of the ensemble members is non-zero.
> >
> > When we compute the rank histogram and there is a "tie" in the
ranks, the
> > code randomly picks a rank from the available ones. There is no
random
> > picking when computing the PHIST. For truly continuous variables,
like
> > temperature, the RHIST and PHIST tell you essentially the same
thing.
> But
> > for discontinuous variables, like precip, the RHIST and PHIST
lines
> treat 0
> > precip differently.
> >
> > When evaluating precip in Ensemble-Stat, it may be wise to only
include
> > observation of precip greater than zero. You'd do this in the
> > Ensemble-Stat config file by setting:
> > obs_thresh = [ >0 ];
> >
> > Thanks,
> > John
> >
> >
> >
> > On Tue, Aug 7, 2018 at 1:36 PM Zhang, Mandy via RT
<met_help at ucar.edu>
> > wrote:
> >
> > >
> > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419 >
> > >
> > > Sorry, my job command is -job aggregate -line_type PHIST
-out_line_type
> > > PHIST -line-fcst_var APCP -fcst_lev A1 -interp_mthd NEAREST,
only
> > > difference is I have -out_line_type PHIST.
> > >
> > > not aggregate_stat
> > >
> > > Mandy
> > >
> > > ________________________________
> > > From: Zhang, Mandy
> > > Sent: Tuesday, August 7, 2018 2:30:47 PM
> > > To: met_help at ucar.edu
> > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> rhist
> > > with STAT_analysis
> > >
> > >
> > > Hi John,
> > >
> > > The methods you told me is what I did before, I use
command:
> > >
> > > -job aggregate_stat -line_type PHIST -out_line_type PHIST
> -line-fcst_var
> > > APCP -fcst_lev A1 -interp_mthd NEAREST, only difference is I
have
> > > -out_line_type PHIST.
> > >
> > > But results of whole period looks wried and the single case
is
> > > reasonable. please find my attached figs.
> > >
> > > And if I want to calculate the CRPS score, does
stat_analysis work
> > > with it?
> > >
> > > Thank you!
> > >
> > > Mandy
> > >
> > > ________________________________
> > > From: John Halley Gotway via RT <met_help at ucar.edu>
> > > Sent: Tuesday, August 7, 2018 2:04:29 PM
> > > To: Zhang, Mandy
> > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> rhist
> > > with STAT_analysis
> > >
> > > Hi Mandy,
> > >
> > > I see you have a question about processing the _orank.nc output
file
> > from
> > > Ensemble-Stat. First, let me briefly review the output lines
type from
> > > Ensemble-Stat...
> > >
> > > The ORANK line type stands for observation rank and contains the
> > > observation value and all of the ensemble member values
corresponding
> to
> > > that observation. Ensemble-Stat can be run to verify against
point
> > > observation and/or gridded observations. When verifying with
point
> > > observations, the ORANK information is written to an ASCII
output file.
> > > But when verifying with gridded observations, the ORANK
information is
> > > written to a NetCDF output file. With gridded obs, there are
usually
> so
> > > many points that it isn't realistic to write them all to ASCII
output.
> > >
> > > The RHIST line type is always written the the ASCII output of
> > Ensemble-Stat
> > > and contains the counts for the binned observation ranks.
Basically,
> > it's
> > > a histogram summarizing the observation ranks across all the
verifying
> > > observation points.
> > >
> > > Next, STAT-Analysis is designed to only process the ASCII output
of
> > > Ensemble-Stat, not the NetCDF file. So STAT-Analysis doesn't
know how
> to
> > > process the _orank.nc files.
> > >
> > > I see that you'd ultimately like to compute a single ranked
histogram
> > over
> > > a whole month of data. There are 2 ways of doing this:
> > > (1) Aggregate the ORANK lines together into one big RHIST
output line
> > > (this is what you're trying but it's not working since you don't
have
> > ORANK
> > > lines in your ASCII output).
> > > (2) Simple aggregate the RHIST lines together over the whole
month.
> > >
> > > Option 2 is much simpler and requires much less data to process.
> Please
> > > try running this job:
> > > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > -line_type RHIST
> > >
> > > And you can try the same thing with the PHIST lines:
> > > -job aggregate -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > -line_type PHIST
> > >
> > > Does that do what you want it to do?
> > >
> > > Thanks,
> > > John
> > >
> > > On Tue, Jul 31, 2018 at 4:35 PM Zhang, Mandy via RT
<met_help at ucar.edu
> >
> > > wrote:
> > >
> > > >
> > > > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
>
> > > >
> > > > Hi Julie
> > > >
> > > > thank you for letting me know!
> > > >
> > > > Mandy
> > > >
> > > > ________________________________
> > > > From: Julie Prestopnik via RT <met_help at ucar.edu>
> > > > Sent: Tuesday, July 31, 2018 4:43:29 PM
> > > > To: Zhang, Mandy
> > > > Subject: Re: [rt.rap.ucar.edu #86419] missing Orank when
calculate
> > rhist
> > > > with STAT_analysis
> > > >
> > > > Hi Mandy. Our MET expert is out of the office this week.
> > > Unfortunately, I
> > > > am unable to help with this issue. I will have our MET expert
follow
> > up
> > > > next week.
> > > >
> > > > Julie
> > > >
> > > > On Tue, Jul 31, 2018 at 3:08 PM Zhang, Mandy via RT <
> met_help at ucar.edu
> > >
> > > > wrote:
> > > >
> > > > >
> > > > > Tue Jul 31 15:07:34 2018: Request 86419 was acted upon.
> > > > > Transaction: Ticket created by Man.Zhang-1 at ou.edu
> > > > > Queue: met_help
> > > > > Subject: missing Orank when calculate rhist with
> STAT_analysis
> > > > > Owner: Nobody
> > > > > Requestors: Man.Zhang-1 at ou.edu
> > > > > Status: new
> > > > > Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=86419
> > > >
> > > > >
> > > > >
> > > > > Zhang, Mandy has shared OneDrive for Business files with
you. To
> view
> > > > > them, click the links below.
> > > > >
> > > > >
> > > > > <
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > > >
> > > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > > >
> > > > >
> > > > > ensemble_stat_ensmr01_20180510_070000V.stat<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ETAmcJEgv7tMnfkJ84So9VUBLw_UXyJeNgDYT0znErRvzw
> > > > > >
> > > > >
> > > > > <
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > > >
> > > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > > >
> > > > >
> > > > > EnsmStatConfig_APCP24<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/EVH0TYZ-ZZBBoAvNMI-ar54BD7Hf-QISxX_KsMDsesME-A
> > > > > >
> > > > >
> > > > > <
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > > >
> > > > > [https://r1.res.office365.com/owa/prem/images/dc-
generic_20.png]<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > > >
> > > > >
> > > > > met_stat_analysis_mon_acc_phist.sh<
> > > > >
> > > >
> > >
> >
> https://sooners-my.sharepoint.com/:u:/g/personal/man_zhang-
1_ou_edu/ESI6WMS-_whAh6bD01QBbdYBO-EXRtVPBB2g9Mk4XGGmiQ
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Hi All,
> > > > >
> > > > > I want to use Stat_analysis to get seasonal mean rank
> > histogram, I
> > > > > find even if I set ORANK = true in ensemble_stat configure
file,
> the
> > > > ORANK
> > > > > can't be found in ensemble**.stat file, and it causes the
error
> > > > information
> > > > > when I run stat_analysis mode:
> > > > >
> > > > > WARNING: do_job_aggr_stat() -> no matching STAT lines found
for
> job:
> > > -job
> > > > > aggregate_stat -fcst_var APCP -fcst_lev A1 -interp_mthd
NEAREST
> > > > -line_type
> > > > > ORANK -dump_row
> > > > > /cnproj1/mzhang/met_test/met/2018/01/job_aggregate_stat.stat
> > > > -out_line_type
> > > > > PHIST -out_bin_size 0.05000
> > > > >
> > > > > *****************************
> > > > >
> > > > > But I have separated files as:
> > > > > ensemble_stat_ensmr01_20180510_060000V_orank.nc,
> > > > >
> > > > > ensemble_stat_ensmr01_20180510_060000V_phist.txt,
> > > > >
> > > > > ensemble_stat_ensmr01_20180510_060000V_rhist.txt...
> > > > >
> > > > > Do you know what's issue of my case?
> > > > >
> > > > > I attached the ensemble_stat configure file,
ensemble_stat**.stat
> and
> > > > > script of aggreate rhist with Stat_analysis.
> > > > >
> > > > > Thank you very much!
> > > > >
> > > > > Mandy
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > >
> > >
> > >
> >
> >
> >
>
>
>
------------------------------------------------
More information about the Met_help
mailing list