Opened 8 years ago

Closed 8 years ago

#295 closed defect (fixed)

Error when monitoring variables containing "_"

Reported by: jgipsl Owned by: somebody
Priority: major Milestone: libIGCM_v2.8.2
Component: PostProcessing Version:
Keywords: Cc:

Description

The monitoring will fail if among the time series following case exist :

Jobname_19800101_19901231_1M_frac_snow.nc
Jobname_19800101_19901231_1M_snow.nc

Making a monitoring on the variable snow will not work :

snow_lands | "snow" | "" | "snow[d=1]" | "Snow Mass (LANDS)" | "kg/m^2" | "Areas[d=1]*Contfrac[d=1]"

Error message :

monitoring01 mesg : File(s) to use :
./CL4.CWRR90msnowSu.3623.L2546m_19800101_19821231_1M_frac_snow.nc
./CL4.CWRR90msnowSu.3623.L2546m_19800101_19821231_1M_snow.nc
monitoring01 mesg : Field to calculate : snow_lands
monitoring01 mesg : file(s) partionned following ./CL4.CWRR90msnowSu.3623.L2546m_19800101_19821231_1M_frac_snow.nc ./CL4.CWRR90msnowSu.3623.L2546m_19800101_19821231_1M_snow.nc
...

monitoring01 mesg : ------------------------------------------------
monitoring01 mesg : part 1 of 2
monitoring01 mesg : cat monitoring01_script_filestopen_11346.jnl
use "./CL4.CWRR90msnowSu.3623.L2546m_19800101_19821231_1M_frac_snow.nc"
monitoring01 mesg : ferret -nojnl -script monitoring01_create_average.jnl '"monitoring01_script_filestopen_11346.jnl" "snow_lands" "snow[d=1]" "Areas[d=1]*Contfrac[d=1]" "Snow Mass (LANDS)" "kg/m^2" "/ccc/scratch/cont003/dsm/p86ghatt/RUN_DIR/5265956_4857/MONITORING/files/SRF_snow_lands_ave.nc_part001.nc" "1"'
monitoring01 error: Check input files and/or operations made on variables

Work-around :

snow_lands | "frac_snow snow" | "" | "snow[d=2]" | "Snow Mass (LANDS)" | "kg/m^2" | "Areas[d=1]*Contfrac[d=1]"

There is still an error message but it seams to work anyway (I think each time snow.nc is copied, also frac_snow.nc tryes to be copied) :

IGCM_sys_Get : error
ln: creating symbolic link `/ccc/scratch/cont003/dsm/p86ghatt/RUN_DIR/5272809_49551/CL5.CWRR90msnowSu.3607.L2546m_19800101_19901231_1M_frac_snow.nc': File exists
ln: creating symbolic link `/ccc/scratch/cont003/dsm/p86ghatt/RUN_DIR/5272809_49551/CL5.CWRR90msnowSu.3607.L2546m_19800101_19901231_1M_frac_snow.nc': File exists
ln: creating symbolic link `/ccc/scratch/cont003/dsm/p86ghatt/RUN_DIR/5272809_49551/CL5.CWRR90msnowSu.3607.L2546m_19800101_19901231_1M_frac_snow.nc': File exists
IGCM_debug_Exit :  IGCM_sys_Get

Change History (3)

comment:1 Changed 8 years ago by sdipsl

  • Milestone set to libIGCM_v2.9

Analyse de Marie Alice ci dessous. L'option 2) me semble être la meilleur.

1) Pour corriger cela, il faut modifier monitoring01 pour trier et garder une seule fois les fichiers à aller chercher.

C’est fait pour l’option -l1 mais pas encore pour l’option -l2.

2) Ou alors on modifie monitoring01 pour aller chercher les fichiers
*_??_%var.nc
au lieu d’aller chercher les fichiers *_%var.nc
En effet, avec la convention IPSL tous les fichiers TS sont de la forme …_1M_var.nc ou…._1M_var.nc ou …._HF_var.nc ou …._3H_var.nc

comment:2 Changed 8 years ago by sdipsl

J'ai testé la solution 2), qui marche bien.

comment:3 Changed 8 years ago by sdipsl

  • Resolution set to fixed
  • Status changed from new to closed

monitoring01 has been modified and deployed

Note: See TracTickets for help on using tickets.