Log of /MITgcm_contrib/verification_other/global_oce_llc90/results/output_adm.txt
Parent Directory
| 
Revision Log
| 
Revision Graph
Revision
1.21 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Wed Mar 8 03:24:18 2017 UTC
(8 years, 4 months ago)
by
gforget
Branch:
MAIN
CVS Tags:
HEAD,
checkpoint66e,
checkpoint66f,
checkpoint66g,
checkpoint66h,
checkpoint66i,
checkpoint66j,
checkpoint66k,
checkpoint66l,
checkpoint66m,
checkpoint66n,
checkpoint66o,
checkpoint67a,
checkpoint67b,
checkpoint67d
Changes since
1.20: +315 -313 lines
Diff to
previous 1.20
- update results following mom_vecinv.F revision which removed adjoint recomputations
Y Y Y Y 16> 7<16 FAIL global_oce_llc90 (e=0, w=52)
Y Y Y Y 16> 5<16 FAIL global_oce_llc90.core2
Y Y Y Y 16> 7<16 FAIL global_oce_llc90.ecco_v4
Y Y Y Y 16> 6<16 FAIL global_oce_llc90.ecmwf
Revision
1.20 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Sat Oct 8 19:46:31 2016 UTC
(8 years, 9 months ago)
by
gforget
Branch:
MAIN
CVS Tags:
checkpoint66a,
checkpoint66b,
checkpoint66c,
checkpoint66d
Changes since
1.19: +331 -300 lines
Diff to
previous 1.19
- update results after adding parenthesis in main diagonal calculation
of implicit vertical diffusion matrix (impldiff.F & gad_implicit_r.F).
- the changes had lead to:
Y Y Y Y 15>16< 5 pass global_oce_cs32 (e=0, w=53)
Y Y Y Y 14>16< 5 pass global_oce_llc90 (e=0, w=53)
Y Y Y Y 8>16< 0 pass global_oce_llc90.ecco_v4
Y Y Y Y 15>16< 5 pass global_oce_llc90.ecmwf
Y Y Y Y>14<13 16 16 16 16 16 16 16 14 13 13 13 14 14 14 16 . . . . . . . . pass global_oce_cs32
Y Y Y Y>10<16 16 16 16 16 16 16 16 13 13 13 13 13 13 13 16 . . . . . . . . pass global_oce_llc90
Y Y Y Y> 9<16 16 12 12 13 16 13 11 13 13 11 12 12 13 11 11 . . . . . . . . FAIL global_oce_llc90.core2
Y Y Y Y> 3<12 14 10 10 10 16 11 9 8 9 8 10 11 11 8 9 . . . . . . . . FAIL global_oce_llc90.ecco_v4
Y Y Y Y> 7<16 13 12 12 12 13 13 10 10 12 10 11 12 12 10 11 . . . . . . . . FAIL global_oce_llc90.ecmwf
Revision
1.18
Fri Oct 23 14:40:32 2015 UTC
(9 years, 9 months ago)
by
gforget
Branch:
MAIN
Changes since
1.17: +0 -0 lines
FILE REMOVED
- move verification experiments results to hidden directory to avoid
running these exps on one core through a plain ./testreport
Revision
1.14 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Fri Oct 10 01:46:37 2014 UTC
(10 years, 9 months ago)
by
gforget
Branch:
MAIN
CVS Tags:
checkpoint65f
Changes since
1.13: +345 -342 lines
Diff to
previous 1.13
- Re-run with latest code AFTER switch to gencost and genctrl.
- Results have not changed except, due to truncation errors :
Y Y Y Y 16> 7< 6 FAIL global_oce_llc90.core2
Y Y Y Y 16> 7< 6 FAIL global_oce_llc90.ecmwf
Revision
1.13 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Fri Oct 10 01:40:48 2014 UTC
(10 years, 9 months ago)
by
gforget
Branch:
MAIN
Changes since
1.12: +671 -352 lines
Diff to
previous 1.12
- Re-run with latest code before switch to gencost and genctrl.
- Results have not changed.
Revision
1.11 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Tue Jul 1 14:13:15 2014 UTC
(11 years ago)
by
gforget
Branch:
MAIN
Changes since
1.10: +635 -566 lines
Diff to
previous 1.10
- update results after viscFacAdj fix (now
viscA4Dfile and viscA4Zfile are actually
used in foward model)
Revision
1.10 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Wed Feb 12 20:10:49 2014 UTC
(11 years, 5 months ago)
by
gforget
Branch:
MAIN
Changes since
1.9: +293 -261 lines
Diff to
previous 1.9
- update results after switching to
#PBS -l select=1:ncpus=16:model=ivy+4:ncpus=20:model=ivy
- results change :
Y Y Y Y 13>16< 4 pass global_oce_llc90 (e=8, w=61)
Y Y Y Y 16>16< 8 pass global_oce_llc90.core2
Y Y Y Y 16>16<16 pass global_oce_llc90.ecco_v4
Y Y Y Y 14>16< 7 pass global_oce_llc90.ecmwf
Revision
1.9 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Fri Jan 10 19:01:27 2014 UTC
(11 years, 6 months ago)
by
gforget
Branch:
MAIN
Changes since
1.8: +160 -138 lines
Diff to
previous 1.8
- update results to checkpoint64q AND with external_forcing_surf.F
recomputation fix checked in on Jan9th2014 (seaice_fake
was executed in AD recomputation for checkpoint64j).
- only forward gradient results change
Y Y Y Y 16>16< 9 pass global_oce_llc90
Y Y Y Y 16>16< 9 pass global_oce_llc90.core2
Y Y Y Y 15>16< 9 pass global_oce_llc90.ecco_v4
Y Y Y Y 16>16< 9 pass global_oce_llc90.ecmwf
Revision
1.8 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Fri Jan 10 18:57:10 2014 UTC
(11 years, 6 months ago)
by
gforget
Branch:
MAIN
Changes since
1.7: +251 -250 lines
Diff to
previous 1.7
- update results to checkpoint64j through july 5th
AND with external_forcing_surf.F
recomputation fix checked in on Jan9th2014 (seaice_fake
was executed in AD recomputation for checkpoint64j).
- results change not due to this issue but due to the
use of 1. rather than drf in ALLOW_BALANCE_FLUXES
which affects truncation :
Y Y Y Y 13>16<16 pass global_oce_llc90
Y Y Y Y 16>16< 6 pass global_oce_llc90.core2
Y Y Y Y 16>16<16 pass global_oce_llc90.ecco_v4
Y Y Y Y 14>16< 5 pass global_oce_llc90.ecmwf
Revision
1.6 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Tue Apr 9 20:06:11 2013 UTC
(12 years, 3 months ago)
by
gforget
Branch:
MAIN
Changes since
1.5: +235 -221 lines
Diff to
previous 1.5
- update libraries
#module load modules comp-intel/11.1.046 mpi/mpt.1.25 netcdf/3.6.0/intel
module load modules comp-intel/2011.2 mpi-sgi/mpt.2.06rp16 netcdf/4.0
- reference result produce with devel queue.
Revision
1.4 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Tue Oct 2 13:59:24 2012 UTC
(12 years, 9 months ago)
by
gforget
Branch:
MAIN
Changes since
1.3: +1161 -1683 lines
Diff to
previous 1.3
- update results after data.exf revision, and added pickup in ecco_v4.
Revision
1.3 -
(
view)
(
download)
(
annotate)
-
[select for diffs]
Sun Sep 30 14:01:27 2012 UTC
(12 years, 9 months ago)
by
gforget
Branch:
MAIN
Changes since
1.2: +1364 -1259 lines
Diff to
previous 1.2
- update results after changes to cal, ctrl, (gmredi,) seaice and data.
- note 1 : the one experiment for which results are unchanged
is ecco_v4. The difference is that the ecco_v4 parameters are
NOT ANY MORE used systematically in run, tr_run.core2, tr_run.ecmwf.
This required the recent change to pkg/gmredi, but code was untouched.
- note 2 : all forward runs but ecco_v4 switched to gkw91 taper scheme
as a result of the recent pkg/gmredi change, where 'stableGmAdjTap'
needs to be set to actually use GMREDI_WITH_STABLE_ADJOINT.
- note 3 : another revision of the experiments is underway that
will involve pkg/ecco changes, the use of core2/cnyf
in main exp, and tweaks to ecco_v4 exp.
This form allows you to request diffs between any two revisions of this file.
For each of the two "sides" of the diff,
select a symbolic revision name using the selection box, or choose
'Use Text Field' and enter a numeric revision.
