Learning to use Dedalus
Dedalus is an open-source PDE solver based on modern spectral methods. It provides a high-level Python interface for creating and solving global spectral discretizations of many PDEs. It is primarily used for computational fluid dynamics (CFD), but can solve many other types of equations as well. This repository gathers tutorials and resources for learning how to use Dedalus for various applications.
Tutorial notebooks
The recommended starting point for learning Dedalus is to go through the official tutorial notebooks in the main code documentation. These describe many of the features of the code, explain the basic class structure, and show typical ways of interacting with Dedalus objects. The examples are designed to run quickly on a single core / laptop.
Example scripts
The main code documentation also contains a range of example scripts showing the typical usage of Dedalus in an HPC environment (i.e. executing model scripts rather than notebooks). These examples are designed to run within several minutes on a single core / laptop, but can be easily scaled up to production sized simulations on a cluster. Many users begin implementing their own models based on one of these scripts.
Practical sessions for GAFD
The following material was originally developed for the practical numerical sessions for the 2023 CISM school on “Fluid Mechanics of Planets and Stars”. These lectures explore how to get started applying Dedalus to problems related to geophysical and astrophysical fluid dynamics. You can open and execute the notebooks online via Google Colab, or download them to run locally after installing Dedalus on your computer.
- Lecture 1: Introduction to Spectral Methods & Dedalus [slides]
- Lecture 2: Basic Dedalus API – Burgers & KdV Equations [colab]
- Lecture 3: Forcing & Analysis – 2D Turbulence [colab]
- Lecture 4: Eigenvalue Problems – Shallow Water & Shear Instability [colab]
- Lecture 5: Boundaries & Tau Terms – Rayleigh-Benard & Spherical Convection [colab]