[Met_help] [rt.rap.ucar.edu #98413] History for Point stat log information request
John Halley Gotway via RT
met_help at ucar.edu
Wed Feb 10 09:08:50 MST 2021
----------------------------------------------------------------
Initial Request
----------------------------------------------------------------
Hello,
Is there anything in point_stat that will inform me on what the latitude/longitude is for the vertices used in the bilinear interpolation during the interpolation? I am currently investigating a problem that I believe to be in the way Python embedding handles field orientation from raw file read in to Point_stat ingestion so I've created an IEEE file containing a "fake" field containing all 0s except for four vertices surrounding an observation location near the center of the grid. Initially, I set the four vertices to be 100 and ran point stat with this file; the resulting MPR file confirmed that the interpolated forecast at this ob location was 100. I then changed the vertices values to the following:
0 0 100 0
X --> X
0 0 0 100
Where X is the observation location and the corners are forecast field point values. Rerunning MET returned a 0 in the interpolated forecast value for the same observation that I got 100 for in the first test run. Therefore, I would like to print out the locations of the points that MET is using during interpolation. Do you have any idea how to do/add that?
Thank you,
Justin
----------------------------------------------------------------
Complete Ticket History
----------------------------------------------------------------
Subject: Point stat log information request
From: John Halley Gotway
Time: Wed Jan 27 17:23:25 2021
Justin,
The existing code certainly doesn't include log messages with that
level of
detail.
But it'd be pretty easy to manually add a log message at the right
spot in
the code, recompile, and re-test. Is that the type of info you're
after? If
so, can you tell me what version of MET you're currently using? I
could
find the right spot to insert that log message into the right file for
that
version of the code... and send you the update file with which to
test.
Sound good?
Thanks,
John Halley Gotway
On Wed, Jan 27, 2021 at 1:41 PM George McCabe via RT
<met_help at ucar.edu>
wrote:
>
> Wed Jan 27 13:40:57 2021: Request 98413 was acted upon.
> Transaction: Given to johnhg (John Halley Gotway) by mccabe
> Queue: met_help
> Subject: Point stat log information request
> Owner: johnhg
> Requestors: justin.tsu at nrlmry.navy.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=98413 >
>
>
> This transaction appears to have no content
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #98413] Point stat log information request
From: Tsu, Mr. Justin
Time: Wed Jan 27 18:13:04 2021
Sounds good. I am using met8.1a. Would it also be easy to add
something that prints out the top and bottom linear interpolation
calculation prioer to the second linear interpolation?
Justin
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Wednesday, January 27, 2021 4:23 PM
To: Tsu, Mr. Justin
Subject: Re: [rt.rap.ucar.edu #98413] Point stat log information
request
Justin,
The existing code certainly doesn't include log messages with that
level of
detail.
But it'd be pretty easy to manually add a log message at the right
spot in
the code, recompile, and re-test. Is that the type of info you're
after? If
so, can you tell me what version of MET you're currently using? I
could
find the right spot to insert that log message into the right file for
that
version of the code... and send you the update file with which to
test.
Sound good?
Thanks,
John Halley Gotway
On Wed, Jan 27, 2021 at 1:41 PM George McCabe via RT
<met_help at ucar.edu>
wrote:
>
> Wed Jan 27 13:40:57 2021: Request 98413 was acted upon.
> Transaction: Given to johnhg (John Halley Gotway) by mccabe
> Queue: met_help
> Subject: Point stat log information request
> Owner: johnhg
> Requestors: justin.tsu at nrlmry.navy.mil
> Status: new
> Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=98413 >
>
>
> This transaction appears to have no content
>
------------------------------------------------
Subject: Point stat log information request
From: John Halley Gotway
Time: Thu Jan 28 13:16:03 2021
Justin,
OK, I created a branch in MET's repo on GitHub named
main_v8.1_interp_bilin
and updated the contents of this file in there:
https://raw.githubusercontent.com/dtcenter/MET/main_v8.1_interp_bilin/met/src/basic/vx_util/interp_util.cc
This link shows you exactly what lines of code I changed in this
version:
https://github.com/dtcenter/MET/commit/5c2c1c5a425befbd22cec4dea16b2179b56fec3e
When I recompile with this change and run Point-Stat at verbosity
level 4,
I see log messages that look like this:
*DEBUG 4: interp_bilin() -> obs(x,y) = (508.12286,256.77464)DEBUG 4:
SW
(508,256) wgt, val = 0.197676, 4.24532DEBUG 4: SE (509,256) wgt, val =
0.0276889, 4.24033DEBUG 4: NW (508,257) wgt, val = 0.679462,
4.07389DEBUG
4: NE (509,257) wgt, val = 0.0951738, 4.23946DEBUG 4: bilin_v =
4.12814*
If you multiple the wgt*val and sum across the 4 points, you should
get the
interpolated bilin_v value. That's also where you can check the
interpolation algorithm to see if it's what you expect.
Thanks,
John
On Wed, Jan 27, 2021 at 6:13 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=98413 >
>
> Sounds good. I am using met8.1a. Would it also be easy to add
something
> that prints out the top and bottom linear interpolation calculation
prioer
> to the second linear interpolation?
>
> Justin
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, January 27, 2021 4:23 PM
> To: Tsu, Mr. Justin
> Subject: Re: [rt.rap.ucar.edu #98413] Point stat log information
request
>
> Justin,
>
> The existing code certainly doesn't include log messages with that
level of
> detail.
>
> But it'd be pretty easy to manually add a log message at the right
spot in
> the code, recompile, and re-test. Is that the type of info you're
after? If
> so, can you tell me what version of MET you're currently using? I
could
> find the right spot to insert that log message into the right file
for that
> version of the code... and send you the update file with which to
test.
>
> Sound good?
>
> Thanks,
> John Halley Gotway
>
> On Wed, Jan 27, 2021 at 1:41 PM George McCabe via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Wed Jan 27 13:40:57 2021: Request 98413 was acted upon.
> > Transaction: Given to johnhg (John Halley Gotway) by mccabe
> > Queue: met_help
> > Subject: Point stat log information request
> > Owner: johnhg
> > Requestors: justin.tsu at nrlmry.navy.mil
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=98413 >
> >
> >
> > This transaction appears to have no content
> >
>
>
>
>
------------------------------------------------
Subject: RE: [rt.rap.ucar.edu #98413] Point stat log information request
From: Tsu, Mr. Justin
Time: Thu Jan 28 15:16:11 2021
Thanks John,
I am getting my IT team to install this patch in our MET module so it
might be a day before I can test. This looks really good though.
Thanks for being so quick with this. I'll get back to you when I run;
hopefully this helps me resolve the differences in how we've been
doing bilin interp in our two systems
Justin
-----Original Message-----
From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
Sent: Thursday, January 28, 2021 12:16 PM
To: Tsu, Mr. Justin
Subject: Re: [rt.rap.ucar.edu #98413] Point stat log information
request
Justin,
OK, I created a branch in MET's repo on GitHub named
main_v8.1_interp_bilin
and updated the contents of this file in there:
https://raw.githubusercontent.com/dtcenter/MET/main_v8.1_interp_bilin/met/src/basic/vx_util/interp_util.cc
This link shows you exactly what lines of code I changed in this
version:
https://github.com/dtcenter/MET/commit/5c2c1c5a425befbd22cec4dea16b2179b56fec3e
When I recompile with this change and run Point-Stat at verbosity
level 4,
I see log messages that look like this:
*DEBUG 4: interp_bilin() -> obs(x,y) = (508.12286,256.77464)DEBUG 4:
SW
(508,256) wgt, val = 0.197676, 4.24532DEBUG 4: SE (509,256) wgt, val =
0.0276889, 4.24033DEBUG 4: NW (508,257) wgt, val = 0.679462,
4.07389DEBUG
4: NE (509,257) wgt, val = 0.0951738, 4.23946DEBUG 4: bilin_v =
4.12814*
If you multiple the wgt*val and sum across the 4 points, you should
get the
interpolated bilin_v value. That's also where you can check the
interpolation algorithm to see if it's what you expect.
Thanks,
John
On Wed, Jan 27, 2021 at 6:13 PM Tsu, Mr. Justin via RT
<met_help at ucar.edu>
wrote:
>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=98413 >
>
> Sounds good. I am using met8.1a. Would it also be easy to add
something
> that prints out the top and bottom linear interpolation calculation
prioer
> to the second linear interpolation?
>
> Justin
>
> -----Original Message-----
> From: John Halley Gotway via RT [mailto:met_help at ucar.edu]
> Sent: Wednesday, January 27, 2021 4:23 PM
> To: Tsu, Mr. Justin
> Subject: Re: [rt.rap.ucar.edu #98413] Point stat log information
request
>
> Justin,
>
> The existing code certainly doesn't include log messages with that
level of
> detail.
>
> But it'd be pretty easy to manually add a log message at the right
spot in
> the code, recompile, and re-test. Is that the type of info you're
after? If
> so, can you tell me what version of MET you're currently using? I
could
> find the right spot to insert that log message into the right file
for that
> version of the code... and send you the update file with which to
test.
>
> Sound good?
>
> Thanks,
> John Halley Gotway
>
> On Wed, Jan 27, 2021 at 1:41 PM George McCabe via RT
<met_help at ucar.edu>
> wrote:
>
> >
> > Wed Jan 27 13:40:57 2021: Request 98413 was acted upon.
> > Transaction: Given to johnhg (John Halley Gotway) by mccabe
> > Queue: met_help
> > Subject: Point stat log information request
> > Owner: johnhg
> > Requestors: justin.tsu at nrlmry.navy.mil
> > Status: new
> > Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=98413 >
> >
> >
> > This transaction appears to have no content
> >
>
>
>
>
------------------------------------------------
More information about the Met_help
mailing list