Lines of Code
[root]/branches/LMDZ5V2.0-dev/libf/dyn3d
| Author | Changes | Lines of Code | Lines per Change |
|---|---|---|---|
| Totals | 22 (100.0%) | 915 (100.0%) | 41.5 |
| lguez | 11 (50.0%) | 690 (75.4%) | 62.7 |
| jghattas | 8 (36.4%) | 203 (22.2%) | 25.3 |
| emillour | 1 (4.5%) | 20 (2.2%) | 20.0 |
| fairhead | 2 (9.1%) | 2 (0.2%) | 1.0 |
Some minor cosmetic improvements (note that a "STOP" should only be followed by an integer; xlf actually considers having a string after a "stop" as a syntax error for a free form input file, and fails to compile).
EM
20 lines of code changed in 1 file:
Conversion to free source form for "inidissip". "comdissipn.h" is now
valid and equivalent for either free source form or fixed source form.
Added compilation files for Gfortran.
224 lines of code changed in 4 files:
Added subroutine grilles_gcm_netcdf_sub containing the first part of program create_fausse_var in file grilles_gcm_netcdf.F . The new subroutine is called in ce0l if parameter grilles_gcm_netcdf=T (default =F). The subroutine creates the file grilles_gcm.nc .
201 lines of code changed in 7 files:
Inclusion de la routine "stats" du modele martien qui permet de sortir
le cycle diurne moyen de différentes variables et l'écart-type.
Pour activer cette routine mettre
callstats=y
dans config.def. Les résultats seront placés dans le fichier stats.nc
Pour rajouter des variables: allez à la fin de physiq.F et prendre modèle sur
les lignes call wstats(...) et ne pas oublier de modifier le nombre de variables
sorties dans statto.h (n2dvar et n2dvar).
Attention: la routine n'est pas optimale (elle stocke ses résultats intermédiaires
dans le fichier stats.nc et les lit et écrit à chaque appel de la physique)
et n'a été testée dans LMDZ pour l'instant que sur PC/gfortran
et sur SX8/monoprocesseur. Encore du travail donc. Mais la routine est utilisée
régulièrement avec le modèle martien :-)
-----------------------
Inclusion of the martian model "stats" routine that outputs the mean diurnal
cycle of variables as well as their standard deviation.
To activate the routine, one needs to put
callstats=y
in the config.def file. Results will be output in a stats.nc file
To add more output variables: go to the end of physiq.F and use the call wstats(...)
lines as a template and do not forget to modify the number of output variables in
statto.h (n2dvar and n3dvar)
Caution: the routine is not optimal (as it stocks preliminary results in the stats.nc
file reading and writing at each physics step) and has only been tested in LMDZ
with a PC/gfortran setup and on a NEC-SX8/monoprocessor setup. Some work then still to
be done. But it is used regularly with the martial model :-)
2 lines of code changed in 2 files:
Put back modification done in revision 1472. This modification was done
to simplify the understanding of the code but it did also slightly change
the results. With this commit we got back the same results as in the revision
before 1472.
2 lines of code changed in 1 file:
Conversion to free source form for "ugeostr". Removed strange "print
*, 301" at the end of "ugeostr".
72 lines of code changed in 3 files:
Conversion to free source form for "disvert" and "iniacademic", no
other change. Bug fix in "fisrtilp": "fraca" could appear in an
expression while not defined.
394 lines of code changed in 4 files: