PySDM v1.2 is out! (+ 3-month student [e-]internship offer)

[slayoo]

Dear PySDM [potential] Users, Contributors, Forkers, Watchers, Stargazers, et al.,
PySDM v1.2 is out: https://github.com/atmos-cloud-sim-uj/PySDM/releases/tag/v1.2

Short intro to start with: PySDM is an open-source Python package implementing the SDM collisional growth Monte-Carlo algorithm of @s-shima. The package features two number-crunching backends based on Numba/LLVM and ThrustRTC/CUDA for multi-threaded CPU and GPU computations, respectively. PySDM code is Pythonic (and fully debuggable with your favourite Python toolset), features decent unit-test coverage, and works out of the box (i.e. with a single pip install ...) on Linux, macOS and Windows. PySDM includes hello-world examples depicting how to start using it from Python, Julia and Matlab. Moreover, example Jupyter notebooks reproducing results from literature are provided and can be easily run in the cloud through such platforms as mybinder.org (CPU only) or Google Colab (both CPU and GPU), clickable badges for launching the example notebooks in the cloud are provided in the project README file. PySDM is developed targetting atmospheric aerosol/cloud simulations and includes representation of processes beyond collisional growth including water vapour condensation/evaporation (using a bespoke adaptive numerical solver) and, as of the current release, enabling to simulate aerosol physical processing by clouds within warm-rain systems. PySDM supports both simple zero-dimensional box/parcel setups, as well as multi-dimensional flow-coupled setups as depicted below (from the left: aerosol concentration, particle mean radius, drizzle concentration; setup as in Arabas et al. 2015):

For more information, see project README file with links to examples and wiki page with links to relevant talks and other materials.

For details on what's new in PySDM v1.2, see the release notes. Among the key new developments, there are: a new timestep-adaptivity and thread-schedulling scheme for coalescence (kudos @piotrbartman) and a new single-column example based on Shipway & Hill 2012 paper.

Disclaimer: PySDM is still in pre-beta stage of development. Use with caution, do expect glitches and reach out to us, so we can help to make it work for you. Feedback and contributions very welcome. Please spread the news and enjoy super droplets!

P.S. We have an open call for two 3-month paid internships aimed at contributing to the effort to couple PySDM with ClimateMachine.jl in a LES setup. This is EU funding granted by the Foundation for Polish Science which has supported PySDM development over the last two years. The internships are offered at the Jagiellonian University, remote work is possible, international collaborators more than welcome! Please point anyone potentially interested to the internship offer page (BSc/MSc student status required for eligibility, application deadline on March 10, internship time: April-June 2021): https://www.fnp.org.pl/en/oferta_pracy/stypendysta/. Please direct any questions to @slayoo.

Greetings,
PySDM Team