[Met_help] [rt.rap.ucar.edu #89165] History for MET ensemble_stat error: no grid defined
John Halley Gotway via RT
met_help at ucar.edu
Thu Mar 14 12:39:43 MDT 2019
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hello,
I am having some trouble running ensemble_stat for WRF forecasts using python embedding and MADIS METAR observations converted to MET format. From my out/error log I see a few problems that I don't understand how to solve. First, the file type I have provided for my observations seems to be ignored, though I specified that I am using a MET configured netCDF file. Second, the level I provide for my observations isn't being read, which may be related to the above. Last, the error I am receiving is that there is no grid defined, which is correct because the observations are not on a grid. Here is my command line entry:
ensemble_stat 1 PYTHON_NUMPY ~/scripts/met/EnsembleStatConfig -point_obs /lustre/work/rmanser/obs/2016042800/metar_2016042800.nc -v 5
I passed a single ensemble member simply to test how ensemble_stat works. I have attached my configuration file, python script, and out/error log for your reference.
Thank you,
- Russell
Russell Manser
Graduate Research Assistant
Atmospheric Science Group
Texas Tech University
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: MET ensemble_stat error: no grid defined
From: John Halley Gotway
Time: Fri Mar 08 14:20:02 2019
Russell,
I have bad news for you. This functionality will not work in met-8.0.
The
python functionality in met-8.0 can only read in a single gridded
field of
data. Ensemble-Stat, in general, requires more than a single field of
data. Instead, for each entry in the ens.field or fcst.field array,
it
needs to read data for each ensemble member.
While that's a logical extension of the current capability, it just
doesn't
exist yet. The python functionality is useful for tools which process
one
field of data a time (like pcp_combine, plot_data_plane, point_stat,
grid_stat, mode, and wavelet_stat). But it is not yet useful for
tools
which require arrays of data (like ensemble_stat, series_analysis, and
mtd).
We do plan to extend this python functionality to...
(1) Read multiple gridded data fields to make it useful for these
tools.
(2) Read point observation data and pass it in memory to Point-Stat
and
Ensemble-Stat.
But in met-8.0, it just won't work.
Thanks,
John Halley Gotway
On Mon, Mar 4, 2019 at 8:59 AM Manser, Russell P via RT
<met_help at ucar.edu>
wrote:
>
> Mon Mar 04 08:59:31 2019: Request 89165 was acted upon.
> Transaction: Ticket created by Russell.P.Manser at ttu.edu
> Queue: met_help
> Subject: MET ensemble_stat error: no grid defined
> Owner: Nobody
> Requestors: Russell.P.Manser at ttu.edu
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=89165 >
>
>
> Hello,
>
>
> I am having some trouble running ensemble_stat for WRF forecasts
using
> python embedding and MADIS METAR observations converted to MET
format. From
> my out/error log I see a few problems that I don't understand how to
solve.
> First, the file type I have provided for my observations seems to be
> ignored, though I specified that I am using a MET configured netCDF
file.
> Second, the level I provide for my observations isn't being read,
which may
> be related to the above. Last, the error I am receiving is that
there is no
> grid defined, which is correct because the observations are not on a
grid.
> Here is my command line entry:
>
>
> ensemble_stat 1 PYTHON_NUMPY ~/scripts/met/EnsembleStatConfig
-point_obs
> /lustre/work/rmanser/obs/2016042800/metar_2016042800.nc -v 5
>
>
> I passed a single ensemble member simply to test how ensemble_stat
works.
> I have attached my configuration file, python script, and out/error
log for
> your reference.
>
>
> Thank you,
>
>
> - Russell
>
>
> Russell Manser
>
> Graduate Research Assistant
>
> Atmospheric Science Group
>
> Texas Tech University
>
>
------------------------------------------------
Subject: MET ensemble_stat error: no grid defined
From: John Halley Gotway
Time: Fri Mar 08 14:26:23 2019
Russell,
I should have mentioned... you could use a temporary workaround. Use
your
Python script to read your data, but then call it using the
pcp_combine
tool's "-add" command. That'll effectively read data using your
Python
script into memory and write it out in a NetCDF format that the other
MET
tools know how to read.
Here's an example, using data from this site:
*https://dtcenter.org/met/users/downloads/analysis_scripts.php
<https://dtcenter.org/met/users/downloads/analysis_scripts.php>*
*wget
https://dtcenter.org/met/users/downloads/python_scripts/read_CPC_binary.py
<https://dtcenter.org/met/users/downloads/python_scripts/read_CPC_binary.py>wget
https://dtcenter.org/met/users/downloads/python_scripts/gefs-legacy-
00z_rfcst-cal_tmean_20170807_8-14day.bin
<https://dtcenter.org/met/users/downloads/python_scripts/gefs-legacy-
00z_rfcst-cal_tmean_20170807_8-14day.bin>*
*/usr/local/met-8.0/bin/pcp_combine -add \*
*PYTHON_NUMPY 'name="read_CPC_binary.py
gefs-legacy-00z_rfcst-cal_tmean_20170807_8-14day.bin 10";' \*
*met.nc <http://met.nc>*
So this is basically using your Python script to reformat data for use
by
MET.
John
On Fri, Mar 8, 2019 at 2:20 PM The RT System itself via RT <
met_help at ucar.edu> wrote:
>
> Fri Mar 08 14:20:03 2019: Request 89165 was acted upon.
> Transaction: Given to johnhg (John Halley Gotway) by RT_System
> Queue: met_help
> Subject: MET ensemble_stat error: no grid defined
> Owner: johnhg
> Requestors: Russell.P.Manser at ttu.edu
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=89165 >
>
>
> This transaction appears to have no content
>
------------------------------------------------
Subject: MET ensemble_stat error: no grid defined
From: Manser, Russell P
Time: Wed Mar 13 15:28:29 2019
Hello John,
I appreciate the workaround! That solved my issues.
Thanks,
- Russell
Russell Manser
Graduate Research Assistant
Atmospheric Science Group
Texas Tech University
________________________________
From: John Halley Gotway via RT <met_help at ucar.edu>
Sent: Friday, March 8, 2019 3:26:23 PM
To: Manser, Russell P
Subject: Re: [rt.rap.ucar.edu #89165] MET ensemble_stat error: no grid
defined
Russell,
I should have mentioned... you could use a temporary workaround. Use
your
Python script to read your data, but then call it using the
pcp_combine
tool's "-add" command. That'll effectively read data using your
Python
script into memory and write it out in a NetCDF format that the other
MET
tools know how to read.
Here's an example, using data from this site:
*https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdtcenter.org%2Fmet%2Fusers%2Fdownloads%2Fanalysis_scripts.php&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871556678&sdata=SZq6Cx2475fdbQE24VhroEvI1Y2Q6T9dfbGMAYEZujo%3D&reserved=0
<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdtcenter.org%2Fmet%2Fusers%2Fdownloads%2Fanalysis_scripts.php&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871556678&sdata=SZq6Cx2475fdbQE24VhroEvI1Y2Q6T9dfbGMAYEZujo%3D&reserved=0>*
*wget
https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdtcenter.org%2Fmet%2Fusers%2Fdownloads%2Fpython_scripts%2Fread_CPC_binary.py&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871556678&sdata=4AdPnibnKOpkJiPKkzChmpviGV1XEQAU2R%2FECqD1jUg%3D&reserved=0
<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdtcenter.org%2Fmet%2Fusers%2Fdownloads%2Fpython_scripts%2Fread_CPC_binary.py&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871556678&sdata=4AdPnibnKOpkJiPKkzChmpviGV1XEQAU2R%2FECqD1jUg%3D&reserved=0>wget
https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdtcenter.org%2Fmet%2Fusers%2Fdownloads%2Fpython_scripts%2Fgefs-
legacy-00z_rfcst-cal_tmean_20170807_8-
14day.bin&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871566683&sdata=0U7IXn1Ba6a0PtTBvsiuKN0ElWZKQ8mxRxqDhNybmRU%3D&reserved=0
<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdtcenter.org%2Fmet%2Fusers%2Fdownloads%2Fpython_scripts%2Fgefs-
legacy-00z_rfcst-cal_tmean_20170807_8-
14day.bin&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871566683&sdata=0U7IXn1Ba6a0PtTBvsiuKN0ElWZKQ8mxRxqDhNybmRU%3D&reserved=0>*
*/usr/local/met-8.0/bin/pcp_combine -add \*
*PYTHON_NUMPY 'name="read_CPC_binary.py
gefs-legacy-00z_rfcst-cal_tmean_20170807_8-14day.bin 10";' \*
*met.nc
<https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fmet.nc&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871566683&sdata=Mp19hSwZjMx3PS6OC%2BrD92MH8EmNjcEGVp%2BvoUbMFAc%3D&reserved=0>*
So this is basically using your Python script to reformat data for use
by
MET.
John
On Fri, Mar 8, 2019 at 2:20 PM The RT System itself via RT <
met_help at ucar.edu> wrote:
>
> Fri Mar 08 14:20:03 2019: Request 89165 was acted upon.
> Transaction: Given to johnhg (John Halley Gotway) by RT_System
> Queue: met_help
> Subject: MET ensemble_stat error: no grid defined
> Owner: johnhg
> Requestors: Russell.P.Manser at ttu.edu
> Status: new
> Ticket <URL:
https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Frt.rap.ucar.edu%2Frt%2FTicket%2FDisplay.html%3Fid%3D89165&data=02%7C01%7CRussell.P.Manser%40ttu.edu%7C1dbc086caa684819eb2108d6a40cb829%7C178a51bf8b2049ffb65556245d5c173c%7C0%7C1%7C636876771871566683&sdata=SvTbFiVJkEDlUIjn%2F7MseU2xnSgyJ4UaEpP9nwA01kM%3D&reserved=0
>
>
>
> This transaction appears to have no content
>
------------------------------------------------
More information about the Met_help
mailing list