[Met_help] [rt.rap.ucar.edu #40071] History for Mode Question

RAL HelpDesk {for Paul Oldenburg} met_help at ucar.edu
Tue Aug 17 08:27:13 MDT 2010


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

Hi all, I would like to use MET Mode to determine the areal coverage of forecast clouds > threshold compared to observed clouds > threshold.  We have developed software based on the R program you sent us earlier to summarize mode obj.txt files.  In this file there is  area of forecast and observed objects and matched forecast and observed objects.  I would assume I should use the matched objects but I would like to confirm.  If there is a better field to use, let me know.

Thanks

Bob Craig


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

Subject: Re: [rt.rap.ucar.edu #40071] Mode Question
From: Paul Oldenburg
Time: Fri Aug 13 12:26:27 2010

Bob,

There are two alternatives.  You can either use the R script that John
sent you, or you can use the mode_analysis tool
in MET.  The following syntax might be useful for your purposes.
Below, I used mode_analysis to compute the sum (among
other statistics) on the mode output generated by the test scripts in
MET.  I think you can use the same syntax on your
mode output files to calculate the sum of all cloud coverage, right?
The -fcst flag will perform the calculation for
the forecast objects only.  Remember to use -obs to get the sum of the
observed object coverage.  Please let me know if
you have any more questions.

Thanks,

Paul


$ bin/mode_analysis -lookin out/mode -summary -simple -single -fcst
-column area

Total mode lines read =  101
Total mode lines kept =   19

Field    N    Min      Max     Mean   StdDev    P10     P25     P50
P75      P90       Sum
-----   --   ----   ------   ------   ------   ----   -----   -----
------   ------   -------
 area   19   4.00   477.00   125.58   133.27   7.60   22.50   85.00
149.50   324.80   2386.00



RAL HelpDesk {for Craig, Robert J Civ USAF AFWA 16 WS/WXN} wrote:
> Fri Aug 13 12:10:55 2010: Request 40071 was acted upon.
> Transaction: Ticket created by craigr at offutt.af.mil
>        Queue: met_help
>      Subject: Mode Question
>        Owner: Nobody
>   Requestors: craigr at offutt.af.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40071 >
>
>
> Hi all, I would like to use MET Mode to determine the areal coverage
of forecast clouds > threshold compared to observed clouds >
threshold.  We have developed software based on the R program you sent
us earlier to summarize mode obj.txt files.  In this file there is
area of forecast and observed objects and matched forecast and
observed objects.  I would assume I should use the matched objects but
I would like to confirm.  If there is a better field to use, let me
know.
>
> Thanks
>
> Bob Craig


------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #40071] Mode Question
From: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Time: Fri Aug 13 12:41:01 2010

Paul, we wrote an IDL based on R script so we summarizing obj.txt
files the same way the R script is.  The question I have is which
fields should I use for area (matched or unmatched)?  

Thanks
Bob
-----Original Message-----
From: RAL HelpDesk {for Paul Oldenburg}
[mailto:met_help at ucar.edu] 
Sent: Friday, August 13, 2010 1:26 PM
To: Craig, Robert J Civ USAF AFWA 16 WS/WXN
Cc: Sittel, Matthew C CTR
USAF AFWA 16 WS/WXN
Subject: Re: [rt.rap.ucar.edu #40071] Mode
Question

Bob,

There are two alternatives.  You can either use
the R script that John sent you, or you can use the mode_analysis tool
in MET.  The following syntax might be useful for your purposes.
Below, I used mode_analysis to compute the sum (among
other
statistics) on the mode output generated by the test scripts in MET.
I think you can use the same syntax on your
mode output files to
calculate the sum of all cloud coverage, right?  The -fcst flag will
perform the calculation for
the forecast objects only.  Remember to
use -obs to get the sum of the observed object coverage.  Please let
me know if
you have any more questions.

Thanks,

Paul


$
bin/mode_analysis -lookin out/mode -summary -simple -single -fcst
-column area

Total mode lines read =  101
Total mode lines kept =
19

Field    N    Min      Max     Mean   StdDev    P10     P25
P50      P75      P90       Sum
-----   --   ----   ------   ------
------   ----   -----   -----   ------   ------   -------
 area   19
4.00   477.00   125.58   133.27   7.60   22.50   85.00   149.50
324.80   2386.00



RAL HelpDesk {for Craig, Robert J Civ USAF
AFWA 16 WS/WXN} wrote:
> Fri Aug 13 12:10:55 2010: Request 40071 was
acted upon.
> Transaction: Ticket created by craigr at offutt.af.mil
>
Queue: met_help
>      Subject: Mode Question
>        Owner: Nobody
>   Requestors: craigr at offutt.af.mil
>       Status: new
>  Ticket
<URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40071 >
> 
>
> Hi all, I would like to use MET Mode to determine the areal coverage
of forecast clouds > threshold compared to observed clouds >
threshold.  We have developed software based on the R program you sent
us earlier to summarize mode obj.txt files.  In this file there is
area of forecast and observed objects and matched forecast and
observed objects.  I would assume I should use the matched objects but
I would like to confirm.  If there is a better field to use, let me
know.
> 
> Thanks
> 
> Bob Craig

------------------------------------------------
Subject: Re: [rt.rap.ucar.edu #40071] Mode Question
From: John Halley Gotway
Time: Fri Aug 13 12:54:35 2010

Hi Bob,

This is John.

It sounds like you want to figure out how the the total forecast
object area compares to the total observed object area.  And that's
certainly a very reasonable question to ask.  To answer it, you'd
look at the total simple object areas of the matched and unmatched
objects COMBINED.  The example mode_analysis command line Paul sent
you computes that total area of the matched and unmatched simple
forecast objects.  And he suggested you run it a second time for the
obs objects.

There are of course a variety of other questions you might want to ask
when analyzing the MODE output.  For example...
- What percent of my forecast objects are matched...      number of
matched simple forecast objects / total number of simple forecast
objects
- What percent of my forecast object areas are matched... area of
matched simple forecast objects / total area simple forecast objects
- The same questions could be asked of the observations.

The sample R-script I sent you (and the MODE-Analysis tool) compute a
lot of these summary metrics.  Here's some sample output of that
Rscript:
   http://www.dtcenter.org/met/users/downloads/Rscripts/mode_summary.out

As the user, you figure out the question you'd like to ask of the MODE
output, and hopefully it can be answered using the summary counts and
areas the Rscript, MODE-Analysis, or your IDL script computes.

Hopefully that helps.  Have a good weekend.

John


RAL HelpDesk {for Craig, Robert J Civ USAF AFWA 16 WS/WXN} wrote:
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40071 >
>
> Paul, we wrote an IDL based on R script so we summarizing obj.txt
files the same way the R script is.  The question I have is which
fields should I use for area (matched or unmatched)?
>
> Thanks
> Bob
>
> -----Original Message-----
> From: RAL HelpDesk {for Paul Oldenburg} [mailto:met_help at ucar.edu]
> Sent: Friday, August 13, 2010 1:26 PM
> To: Craig, Robert J Civ USAF AFWA 16 WS/WXN
> Cc: Sittel, Matthew C CTR USAF AFWA 16 WS/WXN
> Subject: Re: [rt.rap.ucar.edu #40071] Mode Question
>
> Bob,
>
> There are two alternatives.  You can either use the R script that
John sent you, or you can use the mode_analysis tool
> in MET.  The following syntax might be useful for your purposes.
Below, I used mode_analysis to compute the sum (among
> other statistics) on the mode output generated by the test scripts
in MET.  I think you can use the same syntax on your
> mode output files to calculate the sum of all cloud coverage, right?
The -fcst flag will perform the calculation for
> the forecast objects only.  Remember to use -obs to get the sum of
the observed object coverage.  Please let me know if
> you have any more questions.
>
> Thanks,
>
> Paul
>
>
> $ bin/mode_analysis -lookin out/mode -summary -simple -single -fcst
-column area
>
> Total mode lines read =  101
> Total mode lines kept =   19
>
> Field    N    Min      Max     Mean   StdDev    P10     P25     P50
P75      P90       Sum
> -----   --   ----   ------   ------   ------   ----   -----   -----
------   ------   -------
>  area   19   4.00   477.00   125.58   133.27   7.60   22.50   85.00
149.50   324.80   2386.00
>
>
>
> RAL HelpDesk {for Craig, Robert J Civ USAF AFWA 16 WS/WXN} wrote:
>> Fri Aug 13 12:10:55 2010: Request 40071 was acted upon.
>> Transaction: Ticket created by craigr at offutt.af.mil
>>        Queue: met_help
>>      Subject: Mode Question
>>        Owner: Nobody
>>   Requestors: craigr at offutt.af.mil
>>       Status: new
>>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=40071 >
>>
>>
>> Hi all, I would like to use MET Mode to determine the areal
coverage of forecast clouds > threshold compared to observed clouds >
threshold.  We have developed software based on the R program you sent
us earlier to summarize mode obj.txt files.  In this file there is
area of forecast and observed objects and matched forecast and
observed objects.  I would assume I should use the matched objects but
I would like to confirm.  If there is a better field to use, let me
know.
>>
>> Thanks
>>
>> Bob Craig
>
>

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


More information about the Met_help mailing list