Command Line Interface
gcm_toolkit
also provides some functionality to be used from the command line.
The functionalites explained below are scripts, which are automatically installed into your environment, when installing gcm_toolkit
.
You can use them in your terminal by hitting the name of the script.
Information on the required arguments can be found below, or by using the -h
flag.
convert_to_gcmt
The convert_to_gcmt
script is meant to enable an intermediary step between postprocessing/dataanalysis/plotting and simulation outputs.
convert_to_gcmt
reads the raw input of the GCM output and saves it to an intermediary dataformat (netcdf or zarr).
The script could be used in the following way:
Do climate simulation (remotely on cluster)
Run
convert_to_gcmt
to convert the data to netcdfDownload the netcdf data and load it into
GCMT
Work on the data
Usage:
$ convert_to_gcmt -h
usage: convert_to_gcmt [-h] [-c CONFIG]
optional arguments:
-h, --help show this help message and exit
-c CONFIG, --config CONFIG
specify the path to the config file that holds the
config for converting to xarray
The use of a config file (yaml) for the script is optional, but highly recommended, since script will fallback to defaults, which might or might not work for your setup.
If you do not provide any name for a config file, the script will search for convert.yaml
in your current working directory.
The following options are available in the convert.yaml
file:
iters: "all" # list of iterations to load
prefix: ["T","U","V","W"] # The file prefixes that you want to read
tag: "nc_convert" # GCMT tag that should be created for your conversion
gcm: "MITgcm" # The GCM for which we want to load the raw data
load_existing: False # If we want to check for already converted input and just extend if available
data_path: "run" # Path to the raw data
save_path: "results" # Path at which the converted data should be stored
method: "nc" # Output format used for the conversion
update_along_time: False # Only relevant if method == "zarr". Checkout gcm_toolkit.GCMT.save for more info
# (anything else to be passed to gcm_toolkit.GCMT.read_raw)
Note
All of the other arguments are input for gcm_toolkit.GCMT.read_raw()
Checkout gcm_toolkit.GCMT.read_raw()
, gcm_toolkit.GCMT.load()
and gcm_toolkit.GCMT.save()
to understand the usage of the above parameters.