[Met_help] [rt.rap.ucar.edu #46621] History for Test Results for PB2NC bug fix in METV2.0 (UNCLASSIFIED)

[RAL HelpDesk {for John Halley Gotway}]

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

Classification: UNCLASSIFIED
Caveats: NONE

I performed a test to see the effect of the PB2NC bug fix on some
observations in one of my study domains in UT.

Our system admin compiled the METV2.0 with the patches issued in:
"MET_2.0_patches_20110215.gz" which had the bug fix for PB2NC.

I ran PB2NC on PrepBUFR data for 20110420 using the version with the bug fix
and again using my older version without the bug fix.

The output files for each are attached in MS Word documents named as
follows:

"nopatch_PrepBUFR_obs" contains the PB2NC output produced with the older
version.

"PB2NC_patch_PrepBUFR_obs" contains the PB2NC output produced with the
version which has the big fix.

I performed a "diff" between the 2 files and the results are contained in
the attached MS Word doc "difflog".

I counted the instances where data (TMP, RH, DPT and PRMSL) changed and
found the following results:

TMP(GC 11) changed in 18 lines
RH (GC52) changed in 18 lines
DPT (GC17) changed in 6 lines
PRMSL (GC2) changed in 18 lines

Do these results appear normal in view of the intended purpose of the bug
fix made to PB2NC? 

I have not done any testing yet to see the impact of these changes in
Point-Stat results.

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 #46621] Test Results for PB2NC bug fix in  METV2.0 (UNCLASSIFIED)
From: John Halley Gotway
Time: Wed May 04 21:42:54 2011

John,

Sorry for the delay in getting back with you on this.  We're very busy
this week getting ready for a real-time experiment that starts next
week.

I'd say that the differences you're seeing sound consist with the bug
fix.
 If you want to get very precise about it, you could do so using the
"-dump" option for PB2NC for a single PREPBUFR file.  When I was
testing
the bug fix, I used "-dump" to dump the contents of the PREPBUFR file
to
ASCII files.  Please note that dumping to ASCII takes a long time.
The
ASCII output contains the entire event stack for each observation.  I
compared the contents of the event stack to the values PB2NC wrote out
in
the NetCDF file.  And I saw that before the bug fix, PB2NC was writing
out
the values from the bottom of the event stack.  And after the bug fix,
it
was correctly using the top of the event stack.

You're welcome to do something similar if you'd like.

Hope that helps.

John

>
> Mon May 02 12:44:10 2011: Request 46621 was acted upon.
> Transaction: Ticket created by john.raby at us.army.mil
>        Queue: met_help
>      Subject: Test Results for PB2NC bug fix in METV2.0
(UNCLASSIFIED)
>        Owner: Nobody
>   Requestors: john.raby at us.army.mil
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=46621 >
>
>
> Classification: UNCLASSIFIED
> Caveats: NONE
>
> I performed a test to see the effect of the PB2NC bug fix on some
> observations in one of my study domains in UT.
>
> Our system admin compiled the METV2.0 with the patches issued in:
> "MET_2.0_patches_20110215.gz" which had the bug fix for PB2NC.
>
> I ran PB2NC on PrepBUFR data for 20110420 using the version with the
bug
> fix
> and again using my older version without the bug fix.
>
> The output files for each are attached in MS Word documents named as
> follows:
>
> "nopatch_PrepBUFR_obs" contains the PB2NC output produced with the
older
> version.
>
> "PB2NC_patch_PrepBUFR_obs" contains the PB2NC output produced with
the
> version which has the big fix.
>
> I performed a "diff" between the 2 files and the results are
contained in
> the attached MS Word doc "difflog".
>
> I counted the instances where data (TMP, RH, DPT and PRMSL) changed
and
> found the following results:
>
> TMP(GC 11) changed in 18 lines
> RH (GC52) changed in 18 lines
> DPT (GC17) changed in 6 lines
> PRMSL (GC2) changed in 18 lines
>
> Do these results appear normal in view of the intended purpose of
the bug
> fix made to PB2NC?
>
> I have not done any testing yet to see the impact of these changes
in
> Point-Stat results.
>
> 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
>
>
>
>



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