Generic usage¶

Check the help¶

$> COMMAND [SUBCOMMAND] {-h,--help}

Check the version¶

$> COMMAND {-v,--version}

Debug mode¶

Some progress bars informs you about the processing status of the different subcommands. You can switch to a more verbose mode displaying each step with useful additional information.

$> COMMAND {-d,--debug}

Print all results¶

Default is to only print errors, to change this and print all results:

$> COMMAND {-a,--all}

Specify the project¶

The -p/--project argument is used to parse the corresponding configuration INI file. This argument is case-sensitive and has to correspond to a section name of the configuration file(s).

$> nctime SUBCOMMAND {-p,--project} PROJECT_ID

Submit a configuration directory¶

By default, the configuration files are fetched or read from the $ESGINI_DIR environment variable. If not exists, /esg/config/esgcet that is the usual configuration directory on ESGF nodes is used. If you’re checking your data outside of an ESGF node, you have to fetch those project-specific INI files from GitHub using the esgfetchini command line from the esgprep library. Then you can submit another directory to read the configuration files.

$> COMMAND -i /PATH/TO/CONFIG/

Use a logfile¶

Outputs can be logged into a file named PROG-YYYYMMDD-HHMMSS-PID.log. If not, the standard output is used following the verbose mode. By default, the logfiles are stored in a logs folder created in your current working directory (if not exists). It can be changed by adding a optional logfile directory to the flag.

$> COMMAND {-l,--log} [/PATH/TO/LOGDIR/]

Use filters¶

nctime tools will scan your local archive to achieve data quality check. In such a scan, you can filter the file discovery by using a Python regular expression (see re Python library).

The default is to walk through your local filesystem ignoring the hidden folders. It can be change with:

$> COMMAND --ignore-dir PYTHON_REGEX

nctime only considers unhidden NetCDF files by default excuding the regular expression ^\..*$ and including the following one .*\.nc$. It can be independently change with:

$> COMMAND --include-file PYTHON_REGEX --exclude-file PYTHON_REGEX

Keep in mind that --ignore-dir and --exclude-file specify a directory pattern NOT to be matched, while --include-file specifies a filename pattern TO BE matched.

Use multiprocessing¶

nctime uses a multiprocessing interface. This is useful to process a large amount of data, especially in the case of nctxck with the time axis calculation. Set the number of maximal processes to simultaneously treat several files. One process seems sequential processing. Set it -1 to use all available CPU processes (as returned by multiprocessing.cpu_count()). Default is set to 4 processes.

$> COMMAND --max-processes INTEGER

Use libIGCM infos¶

nctime can use your libIGCM info to automatically apply some configuration. This requires to submit the directory including both your config.card and run.card provided by the libIGCM framework:

$> nctime SUBCOMMAND --card /PATH/TO/SUBMISSION/DIRECTORY

Define a reference calendar¶

The reference calendar is the calendar use by nctime to rebuilt theoretical dates during the whole check. By default, the reference calendar is one from the FIRST file scanned. You can specify your own reference calendar with:

$> COMMAND --calendar CALENDAR

Define reference time units¶

The reference time units are use by nctime to rebuilt theoretical dates during the whole check. By default, the reference time units are those from the FIRST file scanned. You can specify your own reference time units with:

$> COMMAND --units "{seconds,minutes,hours,days} since YYYY-MM-DD [HH:mm:ss]"

Overwrites a frequency increment¶

By default, each supported frequency as its own unit and increment (e.g. mon = 1 months). In some case the frequency increment or units can be change, at least for diagnostic purposes. For finer modification, the increment is change for a couple of MIP TABLE and FREQUENCY. The “all” keyword can be used to change the time increment for “all” table or “all” frequencies values.

$> COMMAND --set-inc TABLE:FREQUENCY=INCREMENT[+]UNITS

To change the time increment of sub-hourly files from the CFsubhr table from 30min to 15min:

$> COMMAND --set-inc CFsubhr:subhrPt=15m

To change the time increment of all sub-hourly files whatever the MIP table:

$> COMMAND --set-inc all:subhr=15m

To change the time increment of all CFsubhr files whatever the frequency:

$> COMMAND --set-inc CFsubhr:all=15m

Exit status¶

• Status = 0

  All the files have been successfully processed without errors.

• Status = N

  Errors occur during file scanning and quality checkup. N is the number of errors

• Status = -1

  Argument parsing error.