[Met_help] [rt.rap.ucar.edu #97371] History for Questoin about point_stat config file mask

John Halley Gotway via RT met_help at ucar.edu
Tue Nov 10 09:35:18 MST 2020


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

Hello,

I have a question about the mask dictionary in the point stat configuration file.  In the mask dictionary, I can define "sid" to be a list of stations that I would like to print out stats for.  I can also define "grid" to be "FULL", which according to the documentation calculates stats over the entire domain. I set sid to be a list of 73 stations that I wanted stats for and got my CNT file to output basically 74 lines.... One line for the FULL grid and 73 lines for the specified stations.  The MPR file contains 150 lines.  The break down for these 150 is:

*       1 line for FULL

*       76 lines for each station used in FULL

*       73 lines for each station specified in sid
Note that there are 3 stations specified in sid that are missing from the FULL grid.  My question is, how are stats like RMSE and ME calculated  in FULL compared to a single sid?  Is the FULL RMSE an average of all SIDs used in FULL?

P.S. My RMSE for full is 314.63849 whereas the RSMEs for my individual SIDs range from 0 to 20
P.S.S This is for tau 0, 150 mb geopotential height

Justin


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

Subject: Questoin about point_stat config file mask
From: John Halley Gotway
Time: Mon Nov 09 10:34:09 2020

Hi Justin,

I see you have a question about masking in Point-Stat. You have
described
the matched pair (MPR) output lines generated by Point-Stat.

Yes, FULL uses all points that fall inside the verification domain
being
evaluated. Based on what you wrote, it sounds like FULL consists of 76
matched pairs.
You explicitly listed 73 stations.

If you set, mpr = BOTH, the output "_mpr.txt" should contain 150
lines:
1 header line.
76 lines for the 76 matched pairs which comprise FULL.
73 lines for the 73 individual stations.

You are wondering how continuous statistics, like RMSE and ME,
computed for
the 76 points of FULL versus one point for a single station id.

I would argue that continuous statistics, like RMSE and ME, are
meaningless
when computed over a single pair. ME is simply the MPR:FCST value -
MPR:OBS
value. And it should be exactly equal to RMSE which is computed by
squaring
ME and then taking the square root of that!

Computing continuous stats over 76 pairs makes much more sense. ME is
just
the average of the 76 MPR:FCST - MPR:OBS differences. And RMSE is the
square root of the average squared errors.

If you're interested in writing/analyzing MPR output lines from Point-
Stat,
I'd recommend post-processing them using the Stat-Analysis tool. For
example, perhaps you want to track performance at an individual
station
through time. Run Point-Stat for many times, writing only MPR output
lines.
And then run the following type of Stat-Analysis job:

stat_analysis -lookin point_stat_output_directory -job aggregate_stat
-line_type MPR -out_line_type CNT -by OBS_SID,FCST_VAR, FCST_LEV
-out_stat
mpr_to_cnt.stat

That will...
(1) Read in all the .stat files it recursively finds in the
"point_stat_output_directory" directory.
(2) Sort the pairs into groups for each unique combination of
OBS_SID,FCST_VAR,FCST_LEV values read from the input.
(3) Compute continuous statistics using the MPR's for each group.
(4) Write the resulting statistics to an output file named
mpr_to_cnt.stat.

Then you'd get a meaningful statistic at each station with results
aggregated through time instead of space (like FULL does).

Hope that helps.

John

On Fri, Nov 6, 2020 at 4:29 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:

>
> Fri Nov 06 16:29:36 2020: Request 97371 was acted upon.
> Transaction: Ticket created by justin.tsu at nrlmry.navy.mil
>        Queue: met_help
>      Subject: Questoin about point_stat config file mask
>        Owner: Nobody
>   Requestors: justin.tsu at nrlmry.navy.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
>
>
> Hello,
>
> I have a question about the mask dictionary in the point stat
> configuration file.  In the mask dictionary, I can define "sid" to
be a
> list of stations that I would like to print out stats for.  I can
also
> define "grid" to be "FULL", which according to the documentation
calculates
> stats over the entire domain. I set sid to be a list of 73 stations
that I
> wanted stats for and got my CNT file to output basically 74
lines.... One
> line for the FULL grid and 73 lines for the specified stations.  The
MPR
> file contains 150 lines.  The break down for these 150 is:
>
> *       1 line for FULL
>
> *       76 lines for each station used in FULL
>
> *       73 lines for each station specified in sid
> Note that there are 3 stations specified in sid that are missing
from the
> FULL grid.  My question is, how are stats like RMSE and ME
calculated  in
> FULL compared to a single sid?  Is the FULL RMSE an average of all
SIDs
> used in FULL?
>
> P.S. My RMSE for full is 314.63849 whereas the RSMEs for my
individual
> SIDs range from 0 to 20
> P.S.S This is for tau 0, 150 mb geopotential height
>
> Justin
>
>

------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #97371] Questoin about point_stat config file mask
From: Tsu, Mr. Justin
Time: Mon Nov 09 11:41:03 2020

Thanks John,

That certainly helped to clear things up.  Just for consistency's
sake, I have an excel spreadsheet with the MPR FCST and OB values for
all 76 stations readings the comprise the FULL average, so I went
ahead and tried to compute the average RMSE and ME and did not get the
values found in the CNT file.  In my excel spreadsheet, I got an 76-
Averaged ME and RMSE of 32.1 and 46.5, respectively.  In CNT, it gave
me 32.1 and 314.6, respectively.  So, something is definitely off with
my RMSE.  Just to be clear, for each MPR FCST and ob pair, I
calculated ME by doing MPR FCST - OB then calculated RMSE by doing
SQRT(ME^2).  Then I did an excel columnar average of the ME column and
then the RMSE column.

Why is the CNT RMSE so much higher than what I am getting?

Justin

-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Monday, November 9, 2020 9:34 AM
To: Tsu, Mr. Justin
Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat config
file mask

Hi Justin,

I see you have a question about masking in Point-Stat. You have
described
the matched pair (MPR) output lines generated by Point-Stat.

Yes, FULL uses all points that fall inside the verification domain
being
evaluated. Based on what you wrote, it sounds like FULL consists of 76
matched pairs.
You explicitly listed 73 stations.

If you set, mpr = BOTH, the output "_mpr.txt" should contain 150
lines:
1 header line.
76 lines for the 76 matched pairs which comprise FULL.
73 lines for the 73 individual stations.

You are wondering how continuous statistics, like RMSE and ME,
computed for
the 76 points of FULL versus one point for a single station id.

I would argue that continuous statistics, like RMSE and ME, are
meaningless
when computed over a single pair. ME is simply the MPR:FCST value -
MPR:OBS
value. And it should be exactly equal to RMSE which is computed by
squaring
ME and then taking the square root of that!

Computing continuous stats over 76 pairs makes much more sense. ME is
just
the average of the 76 MPR:FCST - MPR:OBS differences. And RMSE is the
square root of the average squared errors.

If you're interested in writing/analyzing MPR output lines from Point-
Stat,
I'd recommend post-processing them using the Stat-Analysis tool. For
example, perhaps you want to track performance at an individual
station
through time. Run Point-Stat for many times, writing only MPR output
lines.
And then run the following type of Stat-Analysis job:

stat_analysis -lookin point_stat_output_directory -job aggregate_stat
-line_type MPR -out_line_type CNT -by OBS_SID,FCST_VAR, FCST_LEV
-out_stat
mpr_to_cnt.stat

That will...
(1) Read in all the .stat files it recursively finds in the
"point_stat_output_directory" directory.
(2) Sort the pairs into groups for each unique combination of
OBS_SID,FCST_VAR,FCST_LEV values read from the input.
(3) Compute continuous statistics using the MPR's for each group.
(4) Write the resulting statistics to an output file named
mpr_to_cnt.stat.

Then you'd get a meaningful statistic at each station with results
aggregated through time instead of space (like FULL does).

Hope that helps.

John

On Fri, Nov 6, 2020 at 4:29 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:

>
> Fri Nov 06 16:29:36 2020: Request 97371 was acted upon.
> Transaction: Ticket created by justin.tsu at nrlmry.navy.mil
>        Queue: met_help
>      Subject: Questoin about point_stat config file mask
>        Owner: Nobody
>   Requestors: justin.tsu at nrlmry.navy.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
>
>
> Hello,
>
> I have a question about the mask dictionary in the point stat
> configuration file.  In the mask dictionary, I can define "sid" to
be a
> list of stations that I would like to print out stats for.  I can
also
> define "grid" to be "FULL", which according to the documentation
calculates
> stats over the entire domain. I set sid to be a list of 73 stations
that I
> wanted stats for and got my CNT file to output basically 74
lines.... One
> line for the FULL grid and 73 lines for the specified stations.  The
MPR
> file contains 150 lines.  The break down for these 150 is:
>
> *       1 line for FULL
>
> *       76 lines for each station used in FULL
>
> *       73 lines for each station specified in sid
> Note that there are 3 stations specified in sid that are missing
from the
> FULL grid.  My question is, how are stats like RMSE and ME
calculated  in
> FULL compared to a single sid?  Is the FULL RMSE an average of all
SIDs
> used in FULL?
>
> P.S. My RMSE for full is 314.63849 whereas the RSMEs for my
individual
> SIDs range from 0 to 20
> P.S.S This is for tau 0, 150 mb geopotential height
>
> Justin
>
>



------------------------------------------------
Subject: Questoin about point_stat config file mask
From: John Halley Gotway
Time: Mon Nov 09 11:59:20 2020

Justin,

Not sure what's going on. Feel free to send me your excel spreadsheet
and
MPR output file and I'd be happy to take a look.

John

On Mon, Nov 9, 2020 at 11:41 AM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
>
> Thanks John,
>
> That certainly helped to clear things up.  Just for consistency's
sake, I
> have an excel spreadsheet with the MPR FCST and OB values for all 76
> stations readings the comprise the FULL average, so I went ahead and
tried
> to compute the average RMSE and ME and did not get the values found
in the
> CNT file.  In my excel spreadsheet, I got an 76-Averaged ME and RMSE
of
> 32.1 and 46.5, respectively.  In CNT, it gave me 32.1 and 314.6,
> respectively.  So, something is definitely off with my RMSE.  Just
to be
> clear, for each MPR FCST and ob pair, I calculated ME by doing MPR
FCST -
> OB then calculated RMSE by doing SQRT(ME^2).  Then I did an excel
columnar
> average of the ME column and then the RMSE column.
>
> Why is the CNT RMSE so much higher than what I am getting?
>
> Justin
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Monday, November 9, 2020 9:34 AM
> To: Tsu, Mr. Justin
> Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat
config
> file mask
>
> Hi Justin,
>
> I see you have a question about masking in Point-Stat. You have
described
> the matched pair (MPR) output lines generated by Point-Stat.
>
> Yes, FULL uses all points that fall inside the verification domain
being
> evaluated. Based on what you wrote, it sounds like FULL consists of
76
> matched pairs.
> You explicitly listed 73 stations.
>
> If you set, mpr = BOTH, the output "_mpr.txt" should contain 150
lines:
> 1 header line.
> 76 lines for the 76 matched pairs which comprise FULL.
> 73 lines for the 73 individual stations.
>
> You are wondering how continuous statistics, like RMSE and ME,
computed for
> the 76 points of FULL versus one point for a single station id.
>
> I would argue that continuous statistics, like RMSE and ME, are
meaningless
> when computed over a single pair. ME is simply the MPR:FCST value -
MPR:OBS
> value. And it should be exactly equal to RMSE which is computed by
squaring
> ME and then taking the square root of that!
>
> Computing continuous stats over 76 pairs makes much more sense. ME
is just
> the average of the 76 MPR:FCST - MPR:OBS differences. And RMSE is
the
> square root of the average squared errors.
>
> If you're interested in writing/analyzing MPR output lines from
Point-Stat,
> I'd recommend post-processing them using the Stat-Analysis tool. For
> example, perhaps you want to track performance at an individual
station
> through time. Run Point-Stat for many times, writing only MPR output
lines.
> And then run the following type of Stat-Analysis job:
>
> stat_analysis -lookin point_stat_output_directory -job
aggregate_stat
> -line_type MPR -out_line_type CNT -by OBS_SID,FCST_VAR, FCST_LEV
-out_stat
> mpr_to_cnt.stat
>
> That will...
> (1) Read in all the .stat files it recursively finds in the
> "point_stat_output_directory" directory.
> (2) Sort the pairs into groups for each unique combination of
> OBS_SID,FCST_VAR,FCST_LEV values read from the input.
> (3) Compute continuous statistics using the MPR's for each group.
> (4) Write the resulting statistics to an output file named
mpr_to_cnt.stat.
>
> Then you'd get a meaningful statistic at each station with results
> aggregated through time instead of space (like FULL does).
>
> Hope that helps.
>
> John
>
> On Fri, Nov 6, 2020 at 4:29 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Fri Nov 06 16:29:36 2020: Request 97371 was acted upon.
> > Transaction: Ticket created by justin.tsu at nrlmry.navy.mil
> >        Queue: met_help
> >      Subject: Questoin about point_stat config file mask
> >        Owner: Nobody
> >   Requestors: justin.tsu at nrlmry.navy.mil
> >       Status: new
> >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
> >
> >
> > Hello,
> >
> > I have a question about the mask dictionary in the point stat
> > configuration file.  In the mask dictionary, I can define "sid" to
be a
> > list of stations that I would like to print out stats for.  I can
also
> > define "grid" to be "FULL", which according to the documentation
> calculates
> > stats over the entire domain. I set sid to be a list of 73
stations that
> I
> > wanted stats for and got my CNT file to output basically 74
lines.... One
> > line for the FULL grid and 73 lines for the specified stations.
The MPR
> > file contains 150 lines.  The break down for these 150 is:
> >
> > *       1 line for FULL
> >
> > *       76 lines for each station used in FULL
> >
> > *       73 lines for each station specified in sid
> > Note that there are 3 stations specified in sid that are missing
from the
> > FULL grid.  My question is, how are stats like RMSE and ME
calculated  in
> > FULL compared to a single sid?  Is the FULL RMSE an average of all
SIDs
> > used in FULL?
> >
> > P.S. My RMSE for full is 314.63849 whereas the RSMEs for my
individual
> > SIDs range from 0 to 20
> > P.S.S This is for tau 0, 150 mb geopotential height
> >
> > Justin
> >
> >
>
>
>
>

------------------------------------------------
Subject: Questoin about point_stat config file mask
From: Tsu, Mr. Justin
Time: Mon Nov 09 12:08:53 2020

Here are the CNT and MPR files from my point_stast run with no
specified SID.  This will just be the FULL grid with all stations used
in this grid and so there will only be 76 lines in the MPR file.

Justin



-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Monday, November 9, 2020 10:59 AM
To: Tsu, Mr. Justin
Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat config
file mask

Justin,

Not sure what's going on. Feel free to send me your excel spreadsheet
and
MPR output file and I'd be happy to take a look.

John

On Mon, Nov 9, 2020 at 11:41 AM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
>
> Thanks John,
>
> That certainly helped to clear things up.  Just for consistency's
sake, I
> have an excel spreadsheet with the MPR FCST and OB values for all 76
> stations readings the comprise the FULL average, so I went ahead and
tried
> to compute the average RMSE and ME and did not get the values found
in the
> CNT file.  In my excel spreadsheet, I got an 76-Averaged ME and RMSE
of
> 32.1 and 46.5, respectively.  In CNT, it gave me 32.1 and 314.6,
> respectively.  So, something is definitely off with my RMSE.  Just
to be
> clear, for each MPR FCST and ob pair, I calculated ME by doing MPR
FCST -
> OB then calculated RMSE by doing SQRT(ME^2).  Then I did an excel
columnar
> average of the ME column and then the RMSE column.
>
> Why is the CNT RMSE so much higher than what I am getting?
>
> Justin
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Monday, November 9, 2020 9:34 AM
> To: Tsu, Mr. Justin
> Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat
config
> file mask
>
> Hi Justin,
>
> I see you have a question about masking in Point-Stat. You have
described
> the matched pair (MPR) output lines generated by Point-Stat.
>
> Yes, FULL uses all points that fall inside the verification domain
being
> evaluated. Based on what you wrote, it sounds like FULL consists of
76
> matched pairs.
> You explicitly listed 73 stations.
>
> If you set, mpr = BOTH, the output "_mpr.txt" should contain 150
lines:
> 1 header line.
> 76 lines for the 76 matched pairs which comprise FULL.
> 73 lines for the 73 individual stations.
>
> You are wondering how continuous statistics, like RMSE and ME,
computed for
> the 76 points of FULL versus one point for a single station id.
>
> I would argue that continuous statistics, like RMSE and ME, are
meaningless
> when computed over a single pair. ME is simply the MPR:FCST value -
MPR:OBS
> value. And it should be exactly equal to RMSE which is computed by
squaring
> ME and then taking the square root of that!
>
> Computing continuous stats over 76 pairs makes much more sense. ME
is just
> the average of the 76 MPR:FCST - MPR:OBS differences. And RMSE is
the
> square root of the average squared errors.
>
> If you're interested in writing/analyzing MPR output lines from
Point-Stat,
> I'd recommend post-processing them using the Stat-Analysis tool. For
> example, perhaps you want to track performance at an individual
station
> through time. Run Point-Stat for many times, writing only MPR output
lines.
> And then run the following type of Stat-Analysis job:
>
> stat_analysis -lookin point_stat_output_directory -job
aggregate_stat
> -line_type MPR -out_line_type CNT -by OBS_SID,FCST_VAR, FCST_LEV
-out_stat
> mpr_to_cnt.stat
>
> That will...
> (1) Read in all the .stat files it recursively finds in the
> "point_stat_output_directory" directory.
> (2) Sort the pairs into groups for each unique combination of
> OBS_SID,FCST_VAR,FCST_LEV values read from the input.
> (3) Compute continuous statistics using the MPR's for each group.
> (4) Write the resulting statistics to an output file named
mpr_to_cnt.stat.
>
> Then you'd get a meaningful statistic at each station with results
> aggregated through time instead of space (like FULL does).
>
> Hope that helps.
>
> John
>
> On Fri, Nov 6, 2020 at 4:29 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Fri Nov 06 16:29:36 2020: Request 97371 was acted upon.
> > Transaction: Ticket created by justin.tsu at nrlmry.navy.mil
> >        Queue: met_help
> >      Subject: Questoin about point_stat config file mask
> >        Owner: Nobody
> >   Requestors: justin.tsu at nrlmry.navy.mil
> >       Status: new
> >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
> >
> >
> > Hello,
> >
> > I have a question about the mask dictionary in the point stat
> > configuration file.  In the mask dictionary, I can define "sid" to
be a
> > list of stations that I would like to print out stats for.  I can
also
> > define "grid" to be "FULL", which according to the documentation
> calculates
> > stats over the entire domain. I set sid to be a list of 73
stations that
> I
> > wanted stats for and got my CNT file to output basically 74
lines.... One
> > line for the FULL grid and 73 lines for the specified stations.
The MPR
> > file contains 150 lines.  The break down for these 150 is:
> >
> > *       1 line for FULL
> >
> > *       76 lines for each station used in FULL
> >
> > *       73 lines for each station specified in sid
> > Note that there are 3 stations specified in sid that are missing
from the
> > FULL grid.  My question is, how are stats like RMSE and ME
calculated  in
> > FULL compared to a single sid?  Is the FULL RMSE an average of all
SIDs
> > used in FULL?
> >
> > P.S. My RMSE for full is 314.63849 whereas the RSMEs for my
individual
> > SIDs range from 0 to 20
> > P.S.S This is for tau 0, 150 mb geopotential height
> >
> > Justin
> >
> >
>
>
>
>


------------------------------------------------
Subject: Questoin about point_stat config file mask
From: Tsu, Mr. Justin
Time: Mon Nov 09 12:08:53 2020

VERSION MODEL  DESC                 FCST_LEAD FCST_VALID_BEG
FCST_VALID_END  OBS_LEAD OBS_VALID_BEG   OBS_VALID_END   FCST_VAR
FCST_UNITS FCST_LEV   OBS_VAR OBS_UNITS OBS_LEV OBTYPE     VX_MASK
INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA
LINE_TYPE TOTAL FBAR        FBAR_NCL    FBAR_NCU   FBAR_BCL FBAR_BCU
FSTDEV   FSTDEV_NCL FSTDEV_NCU FSTDEV_BCL FSTDEV_BCU OBAR
OBAR_NCL    OBAR_NCU    OBAR_BCL OBAR_BCU OSTDEV    OSTDEV_NCL
OSTDEV_NCU OSTDEV_BCL OSTDEV_BCU PR_CORR PR_CORR_NCL PR_CORR_NCU
PR_CORR_BCL PR_CORR_BCU SP_CORR KT_CORR RANKS FRANK_TIES ORANK_TIES ME
ME_NCL    ME_NCU    ME_BCL ME_BCU ESTDEV   ESTDEV_NCL ESTDEV_NCU
ESTDEV_BCL ESTDEV_BCU MBIAS   MBIAS_BCL MBIAS_BCU MAE      MAE_BCL
MAE_BCU MSE        MSE_BCL MSE_BCU BCMSE       BCMSE_BCL BCMSE_BCU
RMSE      RMSE_BCL RMSE_BCU E10       E10_BCL E10_BCU E25
E25_BCL E25_BCU E50      E50_BCL E50_BCU E75     E75_BCL E75_BCU E90
E90_BCL E90_BCU EIQR     EIQR_BCL EIQR_BCU MAD     MAD_BCL MAD_BCU
ANOM_CORR ANOM_CORR_NCL ANOM_CORR_NCU ANOM_CORR_BCL ANOM_CORR_BCU ME2
ME2_BCL ME2_BCU MSESS    MSESS_BCL MSESS_BCU RMSFA RMSFA_BCL RMSFA_BCU
RMSOA RMSOA_BCL RMSOA_BCU
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190727_223000 20190728_013000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         0.05  CNT          76 14209.51947
14191.44904 14227.5899       NA       NA 80.37613   69.31646
95.67223         NA         NA 14177.40789 14105.08357 14249.73222
NA       NA 321.69411  277.42935  382.91456         NA         NA
0.20643   -0.019957     0.41268          NA          NA      NA
NA     0          0          0 32.11158 -38.72469 102.94784     NA
NA 315.0753  271.72128  375.03615         NA         NA 1.00226
NA        NA 46.51118      NA      NA 98997.3805      NA      NA
97966.22721        NA        NA 314.63849       NA       NA -17.01668
NA      NA -11.38902      NA      NA -7.21806      NA      NA -0.6573
NA      NA 8.18655      NA      NA 10.73173       NA       NA 5.36037
NA      NA        NA            NA            NA            NA
NA 1031.1533      NA      NA 0.043384        NA        NA    NA
NA        NA    NA        NA        NA

------------------------------------------------
Subject: Questoin about point_stat config file mask
From: Tsu, Mr. Justin
Time: Mon Nov 09 12:08:53 2020

VERSION MODEL  DESC                 FCST_LEAD FCST_VALID_BEG
FCST_VALID_END  OBS_LEAD OBS_VALID_BEG   OBS_VALID_END   FCST_VAR
FCST_UNITS FCST_LEV   OBS_VAR OBS_UNITS OBS_LEV OBTYPE     VX_MASK
INTERP_MTHD INTERP_PNTS FCST_THRESH OBS_THRESH COV_THRESH ALPHA
LINE_TYPE TOTAL INDEX OBS_SID OBS_LAT OBS_LON   OBS_LVL OBS_ELV FCST
OBS   OBS_QC CLIMO_MEAN CLIMO_STDEV CLIMO_CDF
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     1   74646
36.6074 262.5119      150   14285 14278.3904  14285   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     2   76394
25.8655 259.75259     150   14278 14272.69681 14278   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     3   76644
20.9469 270.3479      150   14237 14237.59324 14237   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     4   71600
43.9331 299.99301     150   14178 14167.11077 14178   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     5   71603
43.8714 293.89331     150   14129 14114.08378 14129   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     6   71722
46.3019 283.9942      150   14034 14021.39061 14034   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     7   71815
48.5615 301.4342      150   14082 14079.71169 14082   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     8   72201
24.5531 278.2113      150   14239 14236.88228 14239   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76     9   72202
25.75   279.6167      150   11500 14237.36351 11500   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    10   72206
30.4833 278.29889     150   14229 14207.4681  14229   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    11   72210
27.7053 277.59869     150   14245 14230.38243 14245   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    12   72214
30.3931 275.6467      150   14209 14205.07622 14209   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    13   72215
33.3561 275.43271     150   14026 14175.44588 14026   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    14   72230
33.1801 273.21729     150   14175 14167.35294 14175   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    15   72233
30.3376 270.1749      150   14209 14196.45254 14209   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    16   72235
32.32   269.91971     150   14205 14196.28346 14205   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    17   72240
30.1255 266.7829      150   14246 14238.19168 14246   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    18   72248
32.4518 266.15829     150   14255 14251.31314 14255   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    19   72249
32.8351 262.70209     150   14292 14284.67146 14292   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    20   72250
25.9064 262.5744      150   14277 14269.04906 14277   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    21   72251
27.7789 262.4942      150   14281 14272.83722 14281   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    22   72261
29.3745 259.0817      150   14311 14302.16329 14311   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    23   72265
31.9427 257.81021     150   14338 14320.09219 14338   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    24   72274
32.2279 249.04401     150   14349 14348.17333 14349   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    25   72305
34.7758 283.12231     150   14208 14195.76522 14208   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    26   72318
37.2039 279.58578     150   14120 14138.4644  14120   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    27   72327
36.25   273.43329     150   14159 14139.92675 14159   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    28   72340
34.8364 267.7402      150   14225 14222.55049 14225   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    29   72357
35.1808 262.56219     150   14290 14281.93711 14290   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    30   72363
35.2192 258.29419     150   14330 14311.00163 14330   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    31   72364
31.8727 253.3029      150   14346 14336.93002 14346   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    32   72365
35.033  253.383       150   14340 14338.04489 14340   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    33   72402
37.9333 284.51669     150   14170 14157.37583 14170   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    34   72403
38.9667 282.51669     150   14120 14120.24197 14120   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    35   72426
39.4212 276.17889     150   14118 14103.5358  14118   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    36   72440
37.2358 266.59781     150   14241 14230.19297 14241   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    37   72451
37.7616 260.03061     150   14304 14291.74687 14304   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    38   72456
39.0722 264.36938     150   14250 14250.7021  14250   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    39   72469
39.7675 255.1306      150   14310 14301.299   14310   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    40   72476
39.1197 251.47569     150   14319 14328.21414 14319   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    41   72501
40.8651 287.13708     150   14132 14132.20086 14132   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    42   72518
42.6921 286.16739     150   14101 14082.12564 14101   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    43   72520
40.5167 279.78329     150   14108 14086.17647 14108   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    44   72528
42.94   281.2753      150   14064 14052.9185  14064   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    45   72558
41.3195 263.6337      150   14246 14238.73578 14246   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    46   72562
41.1339 259.30011     150   14280 14269.51077 14280   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    47   72632
42.6992 276.52841     150   14082 14078.60364 14082   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    48   72634
44.9083 275.28061     150   14061 14060.85082 14061   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    49   72645
44.4976 271.88831     150   14103 14096.322   14103   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    50   72649
44.849  266.4357      150   14175 14165.48892 14175   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    51   72659
45.4545 261.58578     150   14196 14191.14956 14196   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    52   72672
43.0648 251.5233      150   14255 14264.30246 14255   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    53   72712
46.8682 291.98651     150   14063 14046.87445 14063   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    54   72747
48.5647 266.6026      150   14092 14084.82811 14092   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    55   72764
46.7717 259.2384      150   14170 14168.71981 14170   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    56   72768
48.206  253.3734      150   14136 14136.37625 14136   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    57   74389
43.8925 289.7428      150   14098 14087.25027 14098   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    58   74455
41.6103 269.41171     150   14170 14158.8929  14170   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    59   74560
40.1512 270.66241     150   14165 14152.26156 14165   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    60   74794
28.483  279.45001     150   14240 14227.74238 14240   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    61   76225
28.6707 253.9695      150   14299 14316.03905 14299   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    62   76256
27.9573 249.20261     150   14317 14314.56255 14317   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    63   76458
23.2174 253.589       150   14260 14258.73008 14260   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    64   76612
20.7066 256.6076      150   14258 14263.92797 14258   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    65   76654
19.0443 255.6813      150   14261 14259.93033 14261   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    66   76679
19.433  260.867       150   14260 14267.79097 14260   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    67   76692
19.1425 263.88861     150   14238 14249.31902 14238   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    68   76805
16.7628 260.2507      150   14267 14270.09721 14267   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    69   78073
25.05   282.54001     150   14250 14227.20944 14250   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    70   78384
19.283  278.64999     150   14240 14229.0942  14240   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    71   78486
18.433  290.117       150   14250 14230.76535 14250   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    72   78526
18.43   294           150   14236 14233.36452 14236   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    73   78954
13.0717 300.508       150   14243 14251.58213 14243   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    74   78970
10.617  298.64999     150   14240 14245.10933 14240   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    75   80001
12.583  278.28299     150   14220 14222.89421 14220   None         NA
NA        NA
V8.1    COAMPS CAPE_Nest1a0127x0127 000000    20190728_000000
20190728_000000 000000   20190728_000000 20190728_000000 geopht   gpm
pre_000150 ght     NA        P150    radiosonde FULL    BILIN       4
NA          NA         NA         NA    MPR          76    76   78897
16.25   298.48001     150   14248 14243.62705 14248   None         NA
NA        NA

------------------------------------------------
Subject: Questoin about point_stat config file mask
From: John Halley Gotway
Time: Mon Nov 09 13:31:47 2020

Justin,

OK, sounds good.

(1) Point-Stat reports RMSE = 314.63849
> cat point_stat_CAPE_SINGLE_TEST_000000L_20190728_000000V_cnt.txt |
awk
'{print $78}'
RMSE
314.63849

(2) Stat-Analysis also reports RMSE = 314.6389
>
/Volumes/d1/projects/MET/MET_development/MET-
main_v9.1/met/bin/stat_analysis
-lookin point_stat_CAPE_SINGLE_TEST_000000L_20190728_000000V_mpr.txt
-job
aggregate_stat -line_type MPR -out_line_type CNT | awk '{print $55}'
RMSE
314.63849

The problem is just in your equation for RMSE. Here's the equation for
MSE:
https://dtcenter.github.io/MET/latest/Users_Guide/appendixC.html#mean-
squared-error-mse
Followed by the equation for RMSE:
https://dtcenter.github.io/MET/latest/Users_Guide/appendixC.html#root-
mean-squared-error-rmse

RMSE is the square root of the average of the squared errors for each
point.
The mean of the squared error is 98997.38.
And the square root of that is 314.6385.

See updated spreadsheet attached.

Thanks,
John

On Mon, Nov 9, 2020 at 12:21 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
>
> Here are the CNT and MPR files from my point_stast run with no
specified
> SID.  This will just be the FULL grid with all stations used in this
grid
> and so there will only be 76 lines in the MPR file.
>
> Justin
>
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Monday, November 9, 2020 10:59 AM
> To: Tsu, Mr. Justin
> Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat
config
> file mask
>
> Justin,
>
> Not sure what's going on. Feel free to send me your excel
spreadsheet and
> MPR output file and I'd be happy to take a look.
>
> John
>
> On Mon, Nov 9, 2020 at 11:41 AM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
> >
> > Thanks John,
> >
> > That certainly helped to clear things up.  Just for consistency's
sake, I
> > have an excel spreadsheet with the MPR FCST and OB values for all
76
> > stations readings the comprise the FULL average, so I went ahead
and
> tried
> > to compute the average RMSE and ME and did not get the values
found in
> the
> > CNT file.  In my excel spreadsheet, I got an 76-Averaged ME and
RMSE of
> > 32.1 and 46.5, respectively.  In CNT, it gave me 32.1 and 314.6,
> > respectively.  So, something is definitely off with my RMSE.  Just
to be
> > clear, for each MPR FCST and ob pair, I calculated ME by doing MPR
FCST -
> > OB then calculated RMSE by doing SQRT(ME^2).  Then I did an excel
> columnar
> > average of the ME column and then the RMSE column.
> >
> > Why is the CNT RMSE so much higher than what I am getting?
> >
> > Justin
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Monday, November 9, 2020 9:34 AM
> > To: Tsu, Mr. Justin
> > Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat
config
> > file mask
> >
> > Hi Justin,
> >
> > I see you have a question about masking in Point-Stat. You have
described
> > the matched pair (MPR) output lines generated by Point-Stat.
> >
> > Yes, FULL uses all points that fall inside the verification domain
being
> > evaluated. Based on what you wrote, it sounds like FULL consists
of 76
> > matched pairs.
> > You explicitly listed 73 stations.
> >
> > If you set, mpr = BOTH, the output "_mpr.txt" should contain 150
lines:
> > 1 header line.
> > 76 lines for the 76 matched pairs which comprise FULL.
> > 73 lines for the 73 individual stations.
> >
> > You are wondering how continuous statistics, like RMSE and ME,
computed
> for
> > the 76 points of FULL versus one point for a single station id.
> >
> > I would argue that continuous statistics, like RMSE and ME, are
> meaningless
> > when computed over a single pair. ME is simply the MPR:FCST value
-
> MPR:OBS
> > value. And it should be exactly equal to RMSE which is computed by
> squaring
> > ME and then taking the square root of that!
> >
> > Computing continuous stats over 76 pairs makes much more sense. ME
is
> just
> > the average of the 76 MPR:FCST - MPR:OBS differences. And RMSE is
the
> > square root of the average squared errors.
> >
> > If you're interested in writing/analyzing MPR output lines from
> Point-Stat,
> > I'd recommend post-processing them using the Stat-Analysis tool.
For
> > example, perhaps you want to track performance at an individual
station
> > through time. Run Point-Stat for many times, writing only MPR
output
> lines.
> > And then run the following type of Stat-Analysis job:
> >
> > stat_analysis -lookin point_stat_output_directory -job
aggregate_stat
> > -line_type MPR -out_line_type CNT -by OBS_SID,FCST_VAR, FCST_LEV
> -out_stat
> > mpr_to_cnt.stat
> >
> > That will...
> > (1) Read in all the .stat files it recursively finds in the
> > "point_stat_output_directory" directory.
> > (2) Sort the pairs into groups for each unique combination of
> > OBS_SID,FCST_VAR,FCST_LEV values read from the input.
> > (3) Compute continuous statistics using the MPR's for each group.
> > (4) Write the resulting statistics to an output file named
> mpr_to_cnt.stat.
> >
> > Then you'd get a meaningful statistic at each station with results
> > aggregated through time instead of space (like FULL does).
> >
> > Hope that helps.
> >
> > John
> >
> > On Fri, Nov 6, 2020 at 4:29 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu
> >
> > wrote:
> >
> > >
> > > Fri Nov 06 16:29:36 2020: Request 97371 was acted upon.
> > > Transaction: Ticket created by justin.tsu at nrlmry.navy.mil
> > >        Queue: met_help
> > >      Subject: Questoin about point_stat config file mask
> > >        Owner: Nobody
> > >   Requestors: justin.tsu at nrlmry.navy.mil
> > >       Status: new
> > >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371
> >
> > >
> > >
> > > Hello,
> > >
> > > I have a question about the mask dictionary in the point stat
> > > configuration file.  In the mask dictionary, I can define "sid"
to be a
> > > list of stations that I would like to print out stats for.  I
can also
> > > define "grid" to be "FULL", which according to the documentation
> > calculates
> > > stats over the entire domain. I set sid to be a list of 73
stations
> that
> > I
> > > wanted stats for and got my CNT file to output basically 74
lines....
> One
> > > line for the FULL grid and 73 lines for the specified stations.
The
> MPR
> > > file contains 150 lines.  The break down for these 150 is:
> > >
> > > *       1 line for FULL
> > >
> > > *       76 lines for each station used in FULL
> > >
> > > *       73 lines for each station specified in sid
> > > Note that there are 3 stations specified in sid that are missing
from
> the
> > > FULL grid.  My question is, how are stats like RMSE and ME
calculated
> in
> > > FULL compared to a single sid?  Is the FULL RMSE an average of
all SIDs
> > > used in FULL?
> > >
> > > P.S. My RMSE for full is 314.63849 whereas the RSMEs for my
individual
> > > SIDs range from 0 to 20
> > > P.S.S This is for tau 0, 150 mb geopotential height
> > >
> > > Justin
> > >
> > >
> >
> >
> >
> >
>
>
>

------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #97371] Questoin about point_stat config file mask
From: Tsu, Mr. Justin
Time: Mon Nov 09 17:24:04 2020

Thanks John

This is helpful information

Justin

-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Monday, November 9, 2020 12:32 PM
To: Tsu, Mr. Justin
Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat config
file mask

Justin,

OK, sounds good.

(1) Point-Stat reports RMSE = 314.63849
> cat point_stat_CAPE_SINGLE_TEST_000000L_20190728_000000V_cnt.txt |
awk
'{print $78}'
RMSE
314.63849

(2) Stat-Analysis also reports RMSE = 314.6389
>
/Volumes/d1/projects/MET/MET_development/MET-
main_v9.1/met/bin/stat_analysis
-lookin point_stat_CAPE_SINGLE_TEST_000000L_20190728_000000V_mpr.txt
-job aggregate_stat -line_type MPR -out_line_type CNT | awk '{print
$55}'
RMSE
314.63849

The problem is just in your equation for RMSE. Here's the equation for
MSE:
https://dtcenter.github.io/MET/latest/Users_Guide/appendixC.html#mean-
squared-error-mse
Followed by the equation for RMSE:
https://dtcenter.github.io/MET/latest/Users_Guide/appendixC.html#root-
mean-squared-error-rmse

RMSE is the square root of the average of the squared errors for each
point.
The mean of the squared error is 98997.38.
And the square root of that is 314.6385.

See updated spreadsheet attached.

Thanks,
John

On Mon, Nov 9, 2020 at 12:21 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
>
> Here are the CNT and MPR files from my point_stast run with no
> specified SID.  This will just be the FULL grid with all stations
used
> in this grid and so there will only be 76 lines in the MPR file.
>
> Justin
>
>
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Monday, November 9, 2020 10:59 AM
> To: Tsu, Mr. Justin
> Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat
config
> file mask
>
> Justin,
>
> Not sure what's going on. Feel free to send me your excel
spreadsheet
> and MPR output file and I'd be happy to take a look.
>
> John
>
> On Mon, Nov 9, 2020 at 11:41 AM Tsu, Mr. Justin via RT
> <met_help at ucar.edu>
> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371 >
> >
> > Thanks John,
> >
> > That certainly helped to clear things up.  Just for consistency's
> > sake, I have an excel spreadsheet with the MPR FCST and OB values
> > for all 76 stations readings the comprise the FULL average, so I
> > went ahead and
> tried
> > to compute the average RMSE and ME and did not get the values
found
> > in
> the
> > CNT file.  In my excel spreadsheet, I got an 76-Averaged ME and
RMSE
> > of
> > 32.1 and 46.5, respectively.  In CNT, it gave me 32.1 and 314.6,
> > respectively.  So, something is definitely off with my RMSE.  Just
> > to be clear, for each MPR FCST and ob pair, I calculated ME by
doing
> > MPR FCST - OB then calculated RMSE by doing SQRT(ME^2).  Then I
did
> > an excel
> columnar
> > average of the ME column and then the RMSE column.
> >
> > Why is the CNT RMSE so much higher than what I am getting?
> >
> > Justin
> >
> > -----Original Message-----
> > From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> > Sent: Monday, November 9, 2020 9:34 AM
> > To: Tsu, Mr. Justin
> > Subject: Re: [rt.rap.ucar.edu #97371] Questoin about point_stat
> > config file mask
> >
> > Hi Justin,
> >
> > I see you have a question about masking in Point-Stat. You have
> > described the matched pair (MPR) output lines generated by Point-
Stat.
> >
> > Yes, FULL uses all points that fall inside the verification domain
> > being evaluated. Based on what you wrote, it sounds like FULL
> > consists of 76 matched pairs.
> > You explicitly listed 73 stations.
> >
> > If you set, mpr = BOTH, the output "_mpr.txt" should contain 150
lines:
> > 1 header line.
> > 76 lines for the 76 matched pairs which comprise FULL.
> > 73 lines for the 73 individual stations.
> >
> > You are wondering how continuous statistics, like RMSE and ME,
> > computed
> for
> > the 76 points of FULL versus one point for a single station id.
> >
> > I would argue that continuous statistics, like RMSE and ME, are
> meaningless
> > when computed over a single pair. ME is simply the MPR:FCST value
-
> MPR:OBS
> > value. And it should be exactly equal to RMSE which is computed by
> squaring
> > ME and then taking the square root of that!
> >
> > Computing continuous stats over 76 pairs makes much more sense. ME
> > is
> just
> > the average of the 76 MPR:FCST - MPR:OBS differences. And RMSE is
> > the square root of the average squared errors.
> >
> > If you're interested in writing/analyzing MPR output lines from
> Point-Stat,
> > I'd recommend post-processing them using the Stat-Analysis tool.
For
> > example, perhaps you want to track performance at an individual
> > station through time. Run Point-Stat for many times, writing only
> > MPR output
> lines.
> > And then run the following type of Stat-Analysis job:
> >
> > stat_analysis -lookin point_stat_output_directory -job
> > aggregate_stat -line_type MPR -out_line_type CNT -by
> > OBS_SID,FCST_VAR, FCST_LEV
> -out_stat
> > mpr_to_cnt.stat
> >
> > That will...
> > (1) Read in all the .stat files it recursively finds in the
> > "point_stat_output_directory" directory.
> > (2) Sort the pairs into groups for each unique combination of
> > OBS_SID,FCST_VAR,FCST_LEV values read from the input.
> > (3) Compute continuous statistics using the MPR's for each group.
> > (4) Write the resulting statistics to an output file named
> mpr_to_cnt.stat.
> >
> > Then you'd get a meaningful statistic at each station with results
> > aggregated through time instead of space (like FULL does).
> >
> > Hope that helps.
> >
> > John
> >
> > On Fri, Nov 6, 2020 at 4:29 PM Tsu, Mr. Justin via RT
> > <met_help at ucar.edu
> >
> > wrote:
> >
> > >
> > > Fri Nov 06 16:29:36 2020: Request 97371 was acted upon.
> > > Transaction: Ticket created by justin.tsu at nrlmry.navy.mil
> > >        Queue: met_help
> > >      Subject: Questoin about point_stat config file mask
> > >        Owner: Nobody
> > >   Requestors: justin.tsu at nrlmry.navy.mil
> > >       Status: new
> > >  Ticket <URL:
> > > https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=97371
> >
> > >
> > >
> > > Hello,
> > >
> > > I have a question about the mask dictionary in the point stat
> > > configuration file.  In the mask dictionary, I can define "sid"
to
> > > be a list of stations that I would like to print out stats for.
I
> > > can also define "grid" to be "FULL", which according to the
> > > documentation
> > calculates
> > > stats over the entire domain. I set sid to be a list of 73
> > > stations
> that
> > I
> > > wanted stats for and got my CNT file to output basically 74
lines....
> One
> > > line for the FULL grid and 73 lines for the specified stations.
> > > The
> MPR
> > > file contains 150 lines.  The break down for these 150 is:
> > >
> > > *       1 line for FULL
> > >
> > > *       76 lines for each station used in FULL
> > >
> > > *       73 lines for each station specified in sid
> > > Note that there are 3 stations specified in sid that are missing
> > > from
> the
> > > FULL grid.  My question is, how are stats like RMSE and ME
> > > calculated
> in
> > > FULL compared to a single sid?  Is the FULL RMSE an average of
all
> > > SIDs used in FULL?
> > >
> > > P.S. My RMSE for full is 314.63849 whereas the RSMEs for my
> > > individual SIDs range from 0 to 20 P.S.S This is for tau 0, 150
mb
> > > geopotential height
> > >
> > > Justin
> > >
> > >
> >
> >
> >
> >
>
>
>



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


More information about the Met_help mailing list