The NCL source code for the ILAMB (v1) benchmarking toolkit.
Note that this is not an official ILAMB source repository; it's a just a local copy of CODES_1.4.5 for our permamodel group to experiment with. This is the version of ILAMB that was presented at the 2015 AGU Fall Meeting.
I've written a Python BMI that calls a bash script that runs the NCL version of ILAMB. The run script contains hardcoded paths to the locations of NCL, ImageMagick, and ILAMB on beach, as well as the locations of the CMIP5 and MsTMIP model outputs and benchmark datasets on the data store. So this really only works on beach.
Clone and install into a Python distribution with
$ git clone https://github.com/permamodel/ILAMB
$ cd ILAMB
$ python setup.py install
Setuptools installs the run script ilamb1-run into the bin/ directory of the Python distribution.
In a Python session on beach, execute:
from pylamb import BmiIlamb as ILAMB
x = ILAMB()
x.initialize('/path/to/ILAMB_PARA_SETUP')
x.update() # calls run script
x.finalize()You can provide your own customized ILAMB parameter setup file with either an absolute or a relative path. A sample parameter setup file can be found in the examples/ directory of this repository.
The result:
$ ls -F
ILAMB-output/ ILAMB_PARA_SETUP ILAMB.stderr ILAMB.stdout ILAMB-tmp/Display ILAMB's graphical output with a web browser:
$ firefox ILAMB-output/www/ilamb.html