A lightweight module to store and access magnetic field coefficients for Feltor

The repository serves as a storage for often used magnetic field coefficients for Feltor simulations. Furthermore it provides uniform access to these coefficients through a python interface. Together with the simplesimdb module this interface can then be used to setup and run Feltor simulations that require geometry coefficients on input.

Installation

We do not (yet) have an uploaded version on pypi. Install directly from github:

pip install git+https://github.com/feltor-dev/magneticfielddb

OR clone the repository and then use the package manager pip.

git clone https://github.com/feltor-dev/magneticfielddb
cd magneticfielddb
pip install . # local installation of the module
pip install .[test] # Install pytest
pytest --capture=tee-sys . # run the unittests

In order to locally run the included jupyter-notebooks you can use

pip install .[jupyter] # Install all necessary dependencies
jupyter-lab # run notebooks

Usage

Here is an example of how we iterate over all files in the data repository open them into a python dictionaries and display them on stdout.

### example.py ###
import magneticfielddb as mag

# list all files in the data repository
for f in mag.files() :
    # read each file into a dictionary
    coefficients = mag.select( f)
    # print to stdout
    print( coefficients)

Package Notes

• each file contains at least the "equilibrium", "R_0" and "description" fields
• "R_0" is given in units of meter
• "PP" and "PI" are 1 by default
• Files may contain a field "comment" that contains human readable information string

Additional Resources

• polynomial_field.ipynb is an example notebook of how to fit polynomial coefficients to given equilibrium. Read the doxygen documentation on dg::geo::createMagneticField to find out about valid fields in your json file
• normalize_params.cpp is a C++ program that should be used on geometry input files (all files with X-points must be normalized such that the X-point closest to the O-point lies on the Psip=0 surface). Compile with make and run with ./normalize_params your-params.json your-params.json to update parameters in-place.
• OneSizeFitsAllEquilbrium.nb is a Mathematica notebook that generates solovev coefficients
• q-profiles.ipynb is a jupyter notebook that plots q-profiles and flux surfaces for all equilibria in the database using path/to/feltor/src/geometry_diag/geometry_diag.cpp and simplesimdb
• geometry_diag.ipynb is a jupyter notebook that shows how newly made parameters behave in Feltor and how wall and sheath parameters should be setup for a 3d simulation using path/to/feltor/src/geometry_diag/geometry_diag.cpp and simplesimdb

Contributions

Contributions are welcome.

Authors

Matthias Wiesenberger and Markus Held