Closed svenbuder closed 1 year ago
For the ARE setup, I would add a first warning that you have to load the storage for the catalogs you may want to read in. So maybe just add a one liner: To run through the default tutorial (incl. the loading of "ocean_scalar.*" files from project ik11), you would need access to storage "gdata/tm70+gdata/hh5/+gdata/ik11"
And maybe just add link to "Get started with JupyterLab on NCI": https://opus.nci.org.au/display/OOD/3.1+Starting+JupyterLab+App
Thanks @svenbuder. I'm incorporated this feedback into an overhaul of the docs (#81) that will be merged soon, so I'm closing this issue. Please feel free to add anything else that comes to mind.
Because I have not worked on Gadi before, I was the perfect "new PhD student" test-case and had to take some additional steps to make things work.
We can possibly add them to a "If you have not setup conda on Gadi before, this might be helpful" - Section:
1) Adjust your default project upon login (if you are part of more than 1 project) $ vi $HOME/.config/gadi-login.conf (I had to change mine to tm70, because I accidentally installed into the wrong project directory)
2) If you have not used conda before, initialise your conda $ conda init bash
3) Adjust the commands that are run for conda upon login (note: they will disable automatic environment activations, so that you can activate one of your choice) $ vi ~/.condarc
Once I had done that, I could run your commands without error $ module use /g/data/hh5/public/modules $ module load conda/analysis3 $ conda deactivate $ conda create -n access-nri-catalog -y -c conda-forge -c accessnri jupyterlab dask-labextension matplotlib access-nri-intake