[Met_help] [rt.rap.ucar.edu #46783] History for Problem with METV3.0 (UNCLASSIFIED)

Thu May 12 10:52:27 MDT 2011

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

Classification: UNCLASSIFIED
Caveats: NONE

I performed an experiment to verify some results I got last fall when I
tried to migrate over to using METV3.0 in the midst of calculating
Point-Stat and Stat-Analysis results for some case studies. I reported this
problem to you on 27 OCT 2010. When I compiled and used Stat-Analysis from
METV3.0 to process Point-Stat output which had been previously generated
using METV2.0, I found that the output was nearly all 0-sized files. The log
file of Stat-Analysis activity showed the following error:

"Dataline:: get_item (int) --> range check error" 

This line was written each time it tried to process statistics for 8 surface
parameters. 

The only exception was for WDIR which uses a different line type. In that
case normal statistics were generated.

In this experiment, I was able to reproduce the above error with test data. 

I then re-ran Point-Stat, but this time using METV3.0. In this case
Stat-Analysis (METV3.0) produced good files with normal statistical results.

It appears that METV3.0 is not backward compatible with output generated by
METV2.0. 

I was pleased to find that the statistics I generated from METV3.0
Stat-Analysis were identical with those calculated with METV2.0
Stat-Analysis.

I am anticipating migrating over completely to METV3.0, so I hope to avoid
further issues of this nature, but wanted to see if you can confirm that
this incompatibility is real or if it is possibly due to some other problem.

Thanks.
R/
John

Mr John W. Raby, Meteorologist
U.S. Army Research Laboratory
White Sands Missile Range, NM 88002
(575) 678-2004 DSN 258-2004
FAX (575) 678-1230 DSN 258-1230
Email: john.raby at us.army.mil

Classification: UNCLASSIFIED
Caveats: NONE

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

Subject: Re: [rt.rap.ucar.edu #46783] Problem with METV3.0 (UNCLASSIFIED)
From: Paul Oldenburg
Time: Thu May 12 10:31:01 2011

John,

I ran a stat_analysis filter job and summary job using the METv3.0.1
version of stat_analysis on data generated by
METv2.0 and did not reproduce the problem you reported.  Overall, we
encourage users to use the version of MET that is
consistent with the verification data that they are trying to analyze.
If you would like us to take a closer look at
your error, please send us the stat_analysis job and input data that
you used to generate the error.

Thanks,

Paul

On 05/10/2011 02:24 PM, RAL HelpDesk {for Raby, John} wrote:
>
> Tue May 10 14:24:42 2011: Request 46783 was acted upon.
> Transaction: Ticket created by john.raby at us.army.mil
>        Queue: met_help
>      Subject: Problem with METV3.0 (UNCLASSIFIED)
>        Owner: Nobody
>   Requestors: john.raby at us.army.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46783 >
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
> I performed an experiment to verify some results I got last fall
when I
> tried to migrate over to using METV3.0 in the midst of calculating
> Point-Stat and Stat-Analysis results for some case studies. I
reported this
> problem to you on 27 OCT 2010. When I compiled and used Stat-
Analysis from
> METV3.0 to process Point-Stat output which had been previously
generated
> using METV2.0, I found that the output was nearly all 0-sized files.
The log
> file of Stat-Analysis activity showed the following error:
>
> "Dataline:: get_item (int) --> range check error"
>
> This line was written each time it tried to process statistics for 8
surface
> parameters.
>
> The only exception was for WDIR which uses a different line type. In
that
> case normal statistics were generated.
>
> In this experiment, I was able to reproduce the above error with
test data.
>
> I then re-ran Point-Stat, but this time using METV3.0. In this case
> Stat-Analysis (METV3.0) produced good files with normal statistical
results.
>
> It appears that METV3.0 is not backward compatible with output
generated by
> METV2.0.
>
> I was pleased to find that the statistics I generated from METV3.0
> Stat-Analysis were identical with those calculated with METV2.0
> Stat-Analysis.
>
> I am anticipating migrating over completely to METV3.0, so I hope to
avoid
> further issues of this nature, but wanted to see if you can confirm
that
> this incompatibility is real or if it is possibly due to some other
problem.
>
>
> Thanks.
> R/
> John
>
> Mr John W. Raby, Meteorologist
> U.S. Army Research Laboratory
> White Sands Missile Range, NM 88002
> (575) 678-2004 DSN 258-2004
> FAX (575) 678-1230 DSN 258-1230
> Email: john.raby at us.army.mil
>
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
>
>

------------------------------------------------
Subject: Problem with METV3.0 (UNCLASSIFIED)
From: Raby, John
Time: Thu May 12 10:51:09 2011

Classification: UNCLASSIFIED
Caveats: NONE

Paul -

No need to pursue this further. I have migrated over to METV3.0 now
and so
far, not having any problems. Our local sysadmin has developed a way
for us to
switch versions as needed so I can work with older data. But now I
know that
we must stay consistent and not mix data.

I'm happy to report that METV3.0 includes the fix John Halley Gotway
made to
allow us to handle two sets of observation files from different
sources when
one of them may not have data available for every hour as the other
one does.
MET observations are sometimes sporadic in time with some hours having
data
from a given station and other hours have no observation from that
same
station. In our small domain, this one station was the only one from
one of
our two sources which caused that particular hour to have an empty
file. The
other source did contain usable data for that hour. Before, this would
cause
an error, but after your fix it only issues a warning and moves on to
continue. This is, in our opinion, a very useful feature.

Thanks for your help and advice. This service really helps us.

R/
John

-----Original Message-----
From: RAL HelpDesk {for Paul Oldenburg} [mailto:met_help at ucar.edu]
Sent: Thursday, May 12, 2011 10:31 AM
To: Raby, John (Civ, ARL/CISD)
Cc: Flanigan, Robert (Civ, ARL/CISD)
Subject: Re: [rt.rap.ucar.edu #46783] Problem with METV3.0
(UNCLASSIFIED)

John,

I ran a stat_analysis filter job and summary job using the METv3.0.1
version
of stat_analysis on data generated by METv2.0 and did not reproduce
the
problem you reported.  Overall, we encourage users to use the version
of MET
that is consistent with the verification data that they are trying to
analyze.
If you would like us to take a closer look at your error, please send
us the
stat_analysis job and input data that you used to generate the error.

Thanks,

Paul

On 05/10/2011 02:24 PM, RAL HelpDesk {for Raby, John} wrote:
>
> Tue May 10 14:24:42 2011: Request 46783 was acted upon.
> Transaction: Ticket created by john.raby at us.army.mil
>        Queue: met_help
>      Subject: Problem with METV3.0 (UNCLASSIFIED)
>        Owner: Nobody
>   Requestors: john.raby at us.army.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46783
> >
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
> I performed an experiment to verify some results I got last fall
when
> I tried to migrate over to using METV3.0 in the midst of calculating
> Point-Stat and Stat-Analysis results for some case studies. I
reported
> this problem to you on 27 OCT 2010. When I compiled and used
> Stat-Analysis from METV3.0 to process Point-Stat output which had
been
> previously generated using METV2.0, I found that the output was
nearly
> all 0-sized files. The log file of Stat-Analysis activity showed the
> following error:
>
> "Dataline:: get_item (int) --> range check error"
>
> This line was written each time it tried to process statistics for 8
> surface parameters.
>
> The only exception was for WDIR which uses a different line type. In
> that case normal statistics were generated.
>
> In this experiment, I was able to reproduce the above error with
test data.
>
> I then re-ran Point-Stat, but this time using METV3.0. In this case
> Stat-Analysis (METV3.0) produced good files with normal statistical
results.
>
> It appears that METV3.0 is not backward compatible with output
> generated by METV2.0.
>
> I was pleased to find that the statistics I generated from METV3.0
> Stat-Analysis were identical with those calculated with METV2.0
> Stat-Analysis.
>
> I am anticipating migrating over completely to METV3.0, so I hope to
> avoid further issues of this nature, but wanted to see if you can
> confirm that this incompatibility is real or if it is possibly due
to some
> other problem.
>
>
> Thanks.
> R/
> John
>
> Mr John W. Raby, Meteorologist
> U.S. Army Research Laboratory
> White Sands Missile Range, NM 88002
> (575) 678-2004 DSN 258-2004
> FAX (575) 678-1230 DSN 258-1230
> Email: john.raby at us.army.mil
>
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
>
>

Classification: UNCLASSIFIED
Caveats: NONE

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