ARCCSS Data Access Tools
For full documentation please see http://arccssive.readthedocs.org/en/stable
The stable version of ARCCSSive is available on Rajin in the analysis27
Anaconda environment:
raijin $ module use /g/data3/hh5/public/modules
raijin $ module load conda/analysis27
and is also available as a module:
raijin $ module use ~access/modules
raijin $ module load pythonlib/ARCCSSive
NCI's virtual desktops allow you to use ARCCSSive from a Jupyter notebook. For details on how to use virtual desktops see http://vdi.nci.org.au/help
ARCCSSive can be accessed on VDI using the Anaconda environments:
vdi $ module use /g/data3/hh5/public/modules
vdi $ module load conda/analysis27
You can install ARCCSSive locally using either Anaconda or Pip. You will need to copy the database file from Raijin
$ pip install ARCCSSive
# or
$ conda install -c coecms arccssive
$ scp raijin:/g/data1/ua6/unofficial-ESG-replica/tmp/tree/cmip5_raijin_latest.db $PWD/cmip5.db
$ export CMIP5_DB=sqlite:///$PWD/cmip5.db
To install the current development version with a test database:
$ pip install --user git+https://github.com/coecms/ARCCSSive.git
$ export CMIP5_DB=sqlite:///$HOME/cmip5.db
Query and access the CMIP5 data from Raijin
from ARCCSSive import CMIP5
cmip = CMIP5.DB.connect()
for output in cmip.outputs(model='ACCESS1-0'):
variable = output.variable
files = output.filenames()
Uses SQLAlchemy to filter and sort the data files.