Diagnostics

Setting up Conda Environments and Jupyter Notebooks

To create your conda environment in a terminal load conda module

module load conda

Now create your own conda environment from a pre-existing one

Take a look at this file and examine the included conda packages

less /glade/work/rneale/ASP2023/diags/asp2023_env.yml

Then create your own clone

conda env create -f /glade/work/rneale/ASP2023/diags/asp2023_env.yml

Now load or activate this environment and list the loaded packages

conda activate asp2023
conda list

Set up so jupyterhub can see the asp2023 conda environment as a kernel (say ‘Y’ to any prompts - and then lots of packages will load).

conda install ipykernel
ipython kernel install --user --name=asp2023

If we make changes to our asp2023 environment, run this command each time to update.

conda env update --prefix ./env --file /glade/work/rneale/ASP2023/diags/asp2023_env.yml --prune

You can also create a .yaml file that lists all the packages in your conda environment

conda activate
conda env create -f asp2023_env.yml

Using Python and Jupyter Notebooks

There are a limited number of notebooks available on cheyenne here ls -l /glade/work/rneale/ASP2023/diags

You can compare 2 CESM runs

CESM_2case_comparison.ipynb

You can plot a single level WRF field

wrf_plot_geocat.ipynb

For CESM analysis the CESM tutorial has some simple analysis notebooks

For WRF there is an array of python-based routines to analysis output here

This includes a simple plotting of a skewT plot here

An NCAR-based project GeoCAT has routines that work really well with CESM

For a broader more foundational look at geosciences specific python and jupyter notebooks take a look at Project Pythia