Visualization dashboard for DESC

An interactive application for visualizing three dimensional, toroidal magnetic fields. The dashboard is built in Dash, and is compatible with outputs of the DESC code.

Key features:

• Accepts 'X.h5' files storing Equilibrium or EquilibriumFamily objects from DESC
• Uses DESC's API to compute and store quantities of interest:
  • scalar figures of merit (avg. $\beta$, avg. MHD force error, ...)
  • functions of the flux surface label (pressure, rotational transform, ...)
  • 2D flux surfaces on fixed toroidal slices
  • 2D heatmaps of functions on fixed flux surface or on fixed toroidal slice (current and magnetic field densities, ... )
  • 3D flux surfaces, and the magnetic axis
  • 3D heatmaps of functions on fixed flux surface (current and magnetic field densities, ... )
• At runtime, loads computed quantities into a browser Dash app. User can select different quantities to plot, and cycle through equilibria

Instructions

Step 1. Setup

Clone the repository and set up an appropriate Python environment (eg. using venv). Install the modules desc-opt, dash, and dash-bootstrap-components. Using pip, one can just run

pip install desc-opt dash dash-bootstrap-components

After navigating to the repository in your terminal, you will see /equilibria and /src folders. In /equilibria/desc_files, add the .h5 files from DESC corresponding to your equilibria. Some examples are already included.

Optional

Nagivate to /src/constants.py. There will be a list of the quantities which are precomputed, using notation from DESC's documentation. The list can be modified as desired.

In the same file will be variables base_desc_path and pp_desc_path. These are the paths where the code looks for the .h5 files from DESC, and the preprocessed files, respectively.

Step 2. Preprocessing

Running

python preprocess.py

will scan through the equilibria you have loaded in /equilibria/desc_files (more generally, base_desc_path). The script loads the equilibria, computes any quantities which are needed, and stores the output in a .h5 file in /equilibria/preprocessed (or pp_desc_path).

Other options include

python preprocess.py --list, which lists the equilibria which have been processed.

python preprocess.py --clear, which removes the preprocessed files.

python preprocess.py --clear+, which removes the preprocessed files and then runs the main script.

Step 3. Running the Dash app

The Dash app is now quite simple to run, using

python app.py

The terminal will output "initializing," and list the port from which the app is running. For example, the default is

http://127.0.0.1:8050/

Navigating to this address in a browser should show the app.

Gallery

App running on the W7-X equilibrium, taken from the DESC repo here.

Key Dependencies

This repository relies heavily on the following open source projects:

• DESC (MIT License)
• Dash (MIT License)
• Dash Bootstrap Components (Apache-2.0 License)
• Plotly (MIT License)