Daily Bulletin

JupyterHub kernel and NCAR Python Library updates

May 26, 2022

New versions of the MATLAB (R2022a), R (4.1.2) and Julia (1.7.1) compute kernels are now accessible on NCAR’s JupyterHub service. These match the current defaults found in our Cheyenne and Casper module environments and, in the case of R, the kernel will provide the same library of installed packages. Older versions of these kernels are deprecated and will be removed at a future date (to be announced).

Meanwhile, our NCAR Python Library environment is updated with a new version in a rolling fashion every month. (See Update schedule for conda and the NPL.) For the June release, we will be switching from Python 3.7 to Python 3.8 to provide support for recent releases of Dask components. Python 3.8 provides a number of fundamental language improvements, yet is still supported by a number of legacy packages used at NCAR. This new environment will be available via the conda module on June 1 and will become the default NPL kernel in JupyterHub on July 1.