[Met_help] [rt.rap.ucar.edu #78624] History for point-stat for an ensemble
John Halley Gotway via RT
met_help at ucar.edu
Thu Nov 10 09:15:54 MST 2016
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hi, I'm trying to figure out how best to use MET to compute forecast errors
(f-o) across the full SREF ensemble for four months of data. For each
init/valid time, would I need to run MET independently for each ensemble
member? Or is there a more clever way?
The main thing I need to ultimately generate for my downstream application
is, for all forecasts at a given lead time (or possibly a set of lead
times), a CSV file with forecast errors (f-o) for all ensemble members,
with each line representing a different observation. Is there a
straightforward way to do that with MET/METViewer? Or will I have to write
some of my own scripts to parse the MET output in order to assemble
something like those files?
If you need more clarification or explanation of what I'm trying to do,
please let me know.
Jared
--
===============================
Jared A. Lee, Ph.D.
Project Scientist I
Research Applications Laboratory
National Center for Atmospheric Research
Boulder, Colorado, USA
Email: jaredlee at ucar.edu (w)
Phone: 303.497.8485 (w)
Web: https://staff.ucar.edu/users/jaredlee
===============================
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: point-stat for an ensemble
From: John Halley Gotway
Time: Tue Nov 08 19:22:51 2016
Jared,
Sure, MET can be used to dump out matched pair information for each
ensemble member. You could either do it by processing each ensemble
member
through the Point-Stat tool and dumping out the matched pair line
type.
Alternatively, you could run the Ensemble-Stat tool once for each
init/lead
time and dump out the ORANK line type. That's similar to the MPR line
type, but for each observation value, you get the forecast values for
all
the ensemble members.
My guess is that you'll find Ensemble-Stat most useful and you might
as
well compute ensemble stats, like ranked histograms and roc curves, at
the
same time.
I'll be in my office all day other than noon to 2:30 if you'd like to
come
by and talk through the details.
And we have some perl scripts we've run for ensemble vx in the past
that
you're welcome to adapt for your use.
Thanks,
John Halley Gotway
On Tuesday, November 8, 2016, Jared Lee via RT <met_help at ucar.edu>
wrote:
>
> Tue Nov 08 17:36:38 2016: Request 78624 was acted upon.
> Transaction: Ticket created by jaredlee at ucar.edu <javascript:;>
> Queue: met_help
> Subject: point-stat for an ensemble
> Owner: Nobody
> Requestors: jaredlee at ucar.edu <javascript:;>
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=78624 >
>
>
> Hi, I'm trying to figure out how best to use MET to compute forecast
errors
> (f-o) across the full SREF ensemble for four months of data. For
each
> init/valid time, would I need to run MET independently for each
ensemble
> member? Or is there a more clever way?
>
> The main thing I need to ultimately generate for my downstream
application
> is, for all forecasts at a given lead time (or possibly a set of
lead
> times), a CSV file with forecast errors (f-o) for all ensemble
members,
> with each line representing a different observation. Is there a
> straightforward way to do that with MET/METViewer? Or will I have to
write
> some of my own scripts to parse the MET output in order to assemble
> something like those files?
>
> If you need more clarification or explanation of what I'm trying to
do,
> please let me know.
>
> Jared
>
> --
> ===============================
> Jared A. Lee, Ph.D.
> Project Scientist I
> Research Applications Laboratory
> National Center for Atmospheric Research
> Boulder, Colorado, USA
>
> Email: jaredlee at ucar.edu <javascript:;> (w)
> Phone: 303.497.8485 (w)
> Web: https://staff.ucar.edu/users/jaredlee
> ===============================
>
>
------------------------------------------------
Subject: point-stat for an ensemble
From: Jared Lee
Time: Wed Nov 09 13:03:02 2016
Hi John,
Thanks for the tips. I'll stop by around 2:30 or 3 to chat more about
this.
Jared
On Tue, Nov 8, 2016 at 8:22 PM, John Halley Gotway via RT
<met_help at ucar.edu
> wrote:
> Jared,
>
> Sure, MET can be used to dump out matched pair information for each
> ensemble member. You could either do it by processing each ensemble
member
> through the Point-Stat tool and dumping out the matched pair line
type.
> Alternatively, you could run the Ensemble-Stat tool once for each
init/lead
> time and dump out the ORANK line type. That's similar to the MPR
line
> type, but for each observation value, you get the forecast values
for all
> the ensemble members.
>
> My guess is that you'll find Ensemble-Stat most useful and you might
as
> well compute ensemble stats, like ranked histograms and roc curves,
at the
> same time.
>
> I'll be in my office all day other than noon to 2:30 if you'd like
to come
> by and talk through the details.
>
> And we have some perl scripts we've run for ensemble vx in the past
that
> you're welcome to adapt for your use.
>
> Thanks,
> John Halley Gotway
>
> On Tuesday, November 8, 2016, Jared Lee via RT <met_help at ucar.edu>
wrote:
>
> >
> > Tue Nov 08 17:36:38 2016: Request 78624 was acted upon.
> > Transaction: Ticket created by jaredlee at ucar.edu <javascript:;>
> > Queue: met_help
> > Subject: point-stat for an ensemble
> > Owner: Nobody
> > Requestors: jaredlee at ucar.edu <javascript:;>
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=78624 >
> >
> >
> > Hi, I'm trying to figure out how best to use MET to compute
forecast
> errors
> > (f-o) across the full SREF ensemble for four months of data. For
each
> > init/valid time, would I need to run MET independently for each
ensemble
> > member? Or is there a more clever way?
> >
> > The main thing I need to ultimately generate for my downstream
> application
> > is, for all forecasts at a given lead time (or possibly a set of
lead
> > times), a CSV file with forecast errors (f-o) for all ensemble
members,
> > with each line representing a different observation. Is there a
> > straightforward way to do that with MET/METViewer? Or will I have
to
> write
> > some of my own scripts to parse the MET output in order to
assemble
> > something like those files?
> >
> > If you need more clarification or explanation of what I'm trying
to do,
> > please let me know.
> >
> > Jared
> >
> > --
> > ===============================
> > Jared A. Lee, Ph.D.
> > Project Scientist I
> > Research Applications Laboratory
> > National Center for Atmospheric Research
> > Boulder, Colorado, USA
> >
> > Email: jaredlee at ucar.edu <javascript:;> (w)
> > Phone: 303.497.8485 (w)
> > Web: https://staff.ucar.edu/users/jaredlee
> > ===============================
> >
> >
>
>
--
===============================
Jared A. Lee, Ph.D.
Project Scientist I
Research Applications Laboratory
National Center for Atmospheric Research
Boulder, Colorado, USA
Email: jaredlee at ucar.edu (w)
Phone: 303.497.8485 (w)
Web: https://staff.ucar.edu/users/jaredlee
===============================
------------------------------------------------
Subject: point-stat for an ensemble
From: John Halley Gotway
Time: Wed Nov 09 14:07:03 2016
See you then.
John
On Wed, Nov 9, 2016 at 1:03 PM, Jared Lee via RT <met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=78624 >
>
> Hi John,
>
> Thanks for the tips. I'll stop by around 2:30 or 3 to chat more
about this.
>
> Jared
>
> On Tue, Nov 8, 2016 at 8:22 PM, John Halley Gotway via RT <
> met_help at ucar.edu
> > wrote:
>
> > Jared,
> >
> > Sure, MET can be used to dump out matched pair information for
each
> > ensemble member. You could either do it by processing each
ensemble
> member
> > through the Point-Stat tool and dumping out the matched pair line
type.
> > Alternatively, you could run the Ensemble-Stat tool once for each
> init/lead
> > time and dump out the ORANK line type. That's similar to the MPR
line
> > type, but for each observation value, you get the forecast values
for all
> > the ensemble members.
> >
> > My guess is that you'll find Ensemble-Stat most useful and you
might as
> > well compute ensemble stats, like ranked histograms and roc
curves, at
> the
> > same time.
> >
> > I'll be in my office all day other than noon to 2:30 if you'd like
to
> come
> > by and talk through the details.
> >
> > And we have some perl scripts we've run for ensemble vx in the
past that
> > you're welcome to adapt for your use.
> >
> > Thanks,
> > John Halley Gotway
> >
> > On Tuesday, November 8, 2016, Jared Lee via RT <met_help at ucar.edu>
> wrote:
> >
> > >
> > > Tue Nov 08 17:36:38 2016: Request 78624 was acted upon.
> > > Transaction: Ticket created by jaredlee at ucar.edu <javascript:;>
> > > Queue: met_help
> > > Subject: point-stat for an ensemble
> > > Owner: Nobody
> > > Requestors: jaredlee at ucar.edu <javascript:;>
> > > Status: new
> > > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=78624
> >
> > >
> > >
> > > Hi, I'm trying to figure out how best to use MET to compute
forecast
> > errors
> > > (f-o) across the full SREF ensemble for four months of data. For
each
> > > init/valid time, would I need to run MET independently for each
> ensemble
> > > member? Or is there a more clever way?
> > >
> > > The main thing I need to ultimately generate for my downstream
> > application
> > > is, for all forecasts at a given lead time (or possibly a set of
lead
> > > times), a CSV file with forecast errors (f-o) for all ensemble
members,
> > > with each line representing a different observation. Is there a
> > > straightforward way to do that with MET/METViewer? Or will I
have to
> > write
> > > some of my own scripts to parse the MET output in order to
assemble
> > > something like those files?
> > >
> > > If you need more clarification or explanation of what I'm trying
to do,
> > > please let me know.
> > >
> > > Jared
> > >
> > > --
> > > ===============================
> > > Jared A. Lee, Ph.D.
> > > Project Scientist I
> > > Research Applications Laboratory
> > > National Center for Atmospheric Research
> > > Boulder, Colorado, USA
> > >
> > > Email: jaredlee at ucar.edu <javascript:;> (w)
> > > Phone: 303.497.8485 (w)
> > > Web: https://staff.ucar.edu/users/jaredlee
> > > ===============================
> > >
> > >
> >
> >
>
>
> --
> ===============================
> Jared A. Lee, Ph.D.
> Project Scientist I
> Research Applications Laboratory
> National Center for Atmospheric Research
> Boulder, Colorado, USA
>
> Email: jaredlee at ucar.edu (w)
> Phone: 303.497.8485 (w)
> Web: https://staff.ucar.edu/users/jaredlee
> ===============================
>
>
------------------------------------------------
More information about the Met_help
mailing list