[Met_help] [rt.rap.ucar.edu #48010] History for Stat-Analysis Tool Question

Tue Jul 5 09:45:22 MDT 2011

----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

I'm using stat-analysis to aggregate continuous variables (ME, MSE, RMSE, etc.) for each forecast time.  How does stat-analysis do the aggregation of these variables?  Does it simply just average them for each forecast time?  

Thanks,
Derek Stratman

----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question
From: Paul Oldenburg
Time: Fri Jul 01 12:32:41 2011

Derek,

I assume you are running a job that looks something like this:

  -job aggregate_stat -line_type SL1L2 -out_line_type CNT

In this case, stat_analysis performs the following steps to generate
aggregated statistics:

1.  gather all SL1L2 lines from MET output files in the -lookin
directory
2.  filter the SL1L2 lines based on the criteria that you specified,
like fcst_var = 'TMP'
3.  if you used the -dump_row option, which I highly recommend, all
filtered SL1L2 lines are written to the file you
specified
4.  aggregates the partial sums from the filtered SL1L2 lines and then
computes continuous statistics from the
aggregated partial sums

This process is not the same as averaging the statistics.  This
process was designed to calculate statistics as if all
matched pairs from all valid times (e.g. MET output files) were
grouped together.  Does that make sense?

Paul

On 07/01/2011 10:40 AM, RAL HelpDesk {for Derek.Stratman} wrote:
>
> Fri Jul 01 10:40:38 2011: Request 48010 was acted upon.
> Transaction: Ticket created by Derek.Stratman at noaa.gov
>        Queue: met_help
>      Subject: Stat-Analysis Tool Question
>        Owner: Nobody
>   Requestors: Derek.Stratman at noaa.gov
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=48010 >
>
>
> I'm using stat-analysis to aggregate continuous variables (ME, MSE,
RMSE, etc.) for each forecast time.  How does stat-analysis do the
aggregation of these variables?  Does it simply just average them for
each forecast time?
>
> Thanks,
> Derek Stratman

------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question
From: Derek.Stratman
Time: Tue Jul 05 09:35:48 2011

Paul,

That answered my original question quite well.  One more question
though... what partial sums are FBAR, OBAR, FOBAR, FFBAR, and OOBAR?
I tried searching for the definitions and have a hunch as to what they
are, but I didn't find any documentation.

Thanks,
Derek

----- Original Message -----
From: RAL HelpDesk {for Paul Oldenburg} <met_help at ucar.edu>
Date: Friday, July 1, 2011 1:32 pm
Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question

> Derek,
>
> I assume you are running a job that looks something like this:
>
>  -job aggregate_stat -line_type SL1L2 -out_line_type CNT
>
> In this case, stat_analysis performs the following steps to
> generate aggregated statistics:
>
> 1.  gather all SL1L2 lines from MET output files in the -lookin
> directory2.  filter the SL1L2 lines based on the criteria that you
> specified, like fcst_var = 'TMP'
> 3.  if you used the -dump_row option, which I highly recommend, all
> filtered SL1L2 lines are written to the file you
> specified
> 4.  aggregates the partial sums from the filtered SL1L2 lines and
> then computes continuous statistics from the
> aggregated partial sums
>
> This process is not the same as averaging the statistics.  This
> process was designed to calculate statistics as if all
> matched pairs from all valid times (e.g. MET output files) were
> grouped together.  Does that make sense?
>
> Paul
>
>
>
> On 07/01/2011 10:40 AM, RAL HelpDesk {for Derek.Stratman} wrote:
> >
> > Fri Jul 01 10:40:38 2011: Request 48010 was acted upon.
> > Transaction: Ticket created by Derek.Stratman at noaa.gov
> >        Queue: met_help
> >      Subject: Stat-Analysis Tool Question
> >        Owner: Nobody
> >   Requestors: Derek.Stratman at noaa.gov
> >       Status: new
> >  Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=48010 >
> >
> >
> > I'm using stat-analysis to aggregate continuous variables (ME,
> MSE, RMSE, etc.) for each forecast time.  How does stat-analysis do
> the aggregation of these variables?  Does it simply just average
> them for each forecast time?
> >
> > Thanks,
> > Derek Stratman
>
>
>

------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question
From: Paul Oldenburg
Time: Tue Jul 05 09:44:37 2011

Derek,

See page C-9 in Appendix C of the METv3.0.1 User's Guide.  The section
called _Scalar L1 and L2 values_ describes those
quantities.

Paul

On 07/05/2011 09:35 AM, RAL HelpDesk {for Derek.Stratman} wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=48010 >
>
> Paul,
>
> That answered my original question quite well.  One more question
though... what partial sums are FBAR, OBAR, FOBAR, FFBAR, and OOBAR?
I tried searching for the definitions and have a hunch as to what they
are, but I didn't find any documentation.
>
> Thanks,
> Derek
>
>
> ----- Original Message -----
> From: RAL HelpDesk {for Paul Oldenburg} <met_help at ucar.edu>
> Date: Friday, July 1, 2011 1:32 pm
> Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question
>
>> Derek,
>>
>> I assume you are running a job that looks something like this:
>>
>>  -job aggregate_stat -line_type SL1L2 -out_line_type CNT
>>
>> In this case, stat_analysis performs the following steps to
>> generate aggregated statistics:
>>
>> 1.  gather all SL1L2 lines from MET output files in the -lookin
>> directory2.  filter the SL1L2 lines based on the criteria that you
>> specified, like fcst_var = 'TMP'
>> 3.  if you used the -dump_row option, which I highly recommend, all
>> filtered SL1L2 lines are written to the file you
>> specified
>> 4.  aggregates the partial sums from the filtered SL1L2 lines and
>> then computes continuous statistics from the
>> aggregated partial sums
>>
>> This process is not the same as averaging the statistics.  This
>> process was designed to calculate statistics as if all
>> matched pairs from all valid times (e.g. MET output files) were
>> grouped together.  Does that make sense?
>>
>> Paul
>>
>>
>>
>> On 07/01/2011 10:40 AM, RAL HelpDesk {for Derek.Stratman} wrote:
>>>
>>> Fri Jul 01 10:40:38 2011: Request 48010 was acted upon.
>>> Transaction: Ticket created by Derek.Stratman at noaa.gov
>>>        Queue: met_help
>>>      Subject: Stat-Analysis Tool Question
>>>        Owner: Nobody
>>>   Requestors: Derek.Stratman at noaa.gov
>>>       Status: new
>>>  Ticket <URL:
>> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=48010 >
>>>
>>>
>>> I'm using stat-analysis to aggregate continuous variables (ME,
>> MSE, RMSE, etc.) for each forecast time.  How does stat-analysis do
>> the aggregation of these variables?  Does it simply just average
>> them for each forecast time?
>>>
>>> Thanks,
>>> Derek Stratman
>>
>>
>>

------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question
From: Derek.Stratman
Time: Tue Jul 05 09:44:58 2011

Paul,

Of course I would find the documentation after I emailed you!  I don't
think I have any further questions on this matter now.

Thanks for you help,
Derek

----- Original Message -----
From: "RAL HelpDesk {for Derek.Stratman}" <met_help at ucar.edu>
Date: Tuesday, July 5, 2011 10:35 am
Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question

> Paul,
>
> That answered my original question quite well.  One more question
> though... what partial sums are FBAR, OBAR, FOBAR, FFBAR, and
> OOBAR?  I tried searching for the definitions and have a hunch as
> to what they are, but I didn't find any documentation.
>
> Thanks,
> Derek
>
>
> ----- Original Message -----
> From: RAL HelpDesk {for Paul Oldenburg} <met_help at ucar.edu>
> Date: Friday, July 1, 2011 1:32 pm
> Subject: Re: [rt.rap.ucar.edu #48010] Stat-Analysis Tool Question
>
> > Derek,
> >
> > I assume you are running a job that looks something like this:
> >
> >  -job aggregate_stat -line_type SL1L2 -out_line_type CNT
> >
> > In this case, stat_analysis performs the following steps to
> > generate aggregated statistics:
> >
> > 1.  gather all SL1L2 lines from MET output files in the -lookin
> > directory2.  filter the SL1L2 lines based on the criteria that
> you
> > specified, like fcst_var = 'TMP'
> > 3.  if you used the -dump_row option, which I highly recommend,
> all
> > filtered SL1L2 lines are written to the file you
> > specified
> > 4.  aggregates the partial sums from the filtered SL1L2 lines and
> > then computes continuous statistics from the
> > aggregated partial sums
> >
> > This process is not the same as averaging the statistics.  This
> > process was designed to calculate statistics as if all
> > matched pairs from all valid times (e.g. MET output files) were
> > grouped together.  Does that make sense?
> >
> > Paul
> >
> >
> >
> > On 07/01/2011 10:40 AM, RAL HelpDesk {for Derek.Stratman} wrote:
> > >
> > > Fri Jul 01 10:40:38 2011: Request 48010 was acted upon.
> > > Transaction: Ticket created by Derek.Stratman at noaa.gov
> > >        Queue: met_help
> > >      Subject: Stat-Analysis Tool Question
> > >        Owner: Nobody
> > >   Requestors: Derek.Stratman at noaa.gov
> > >       Status: new
> > >  Ticket <URL:
> > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=48010 >
> > >
> > >
> > > I'm using stat-analysis to aggregate continuous variables (ME,
> > MSE, RMSE, etc.) for each forecast time.  How does stat-analysis
> do
> > the aggregation of these variables?  Does it simply just average
> > them for each forecast time?
> > >
> > > Thanks,
> > > Derek Stratman
> >
> >
> >
>
>

------------------------------------------------