The Generic Interface for MPI-AMRVAC / Legolas Interconnection, or GIMLI for short, is a Python framework to generate user modules and parfiles for both Legolas and the non-linear simulation code MPI-AMRVAC, and to facilitate the use of one code’s output as input in the other. Concretely, GIMLI consists of three parts:

  1. Legolas setup. Relying on Python’s symbolic package SymPy, GIMLI allows you to define a plasma equilibrium analytically with the parameters available in Legolas. Passing this equilibrium along with a dictionary of parameter values and code configuration input such as the solver and which data to save, GIMLI will calculate all the equilibrium’s derivatives for you and generate a Legolas user module smod_user_defined.f08 and accompanying parfile(s). After compilation, it suffices to run Legolas with the generated parfile(s).

  2. MPI-AMRVAC setup. From the same SymPy-based equilibrium definition, with the appropriate accompanying code configuration details, GIMLI can also generate a user module mod_usr.t and parfile for MPI-AMRVAC (v3.3+). The perturbation of this equilibrium is provided by selecting eigenmodes from a Legolas spectrum to construct a superposition of linear solutions with a specified amplitude. This linear solution is then interpolated to the grid in MPI-AMRVAC.

  3. Numerical data in Legolas. With the introduction of GIMLI, Legolas was extended with the option to define equilibria numerically (see numerical equilibrium). GIMLI provides the functionalities to transform numerical arrays, e.g. one-dimensional slices from MPI-AMRVAC data, to a formatted file that can be read by Legolas. Note though that this requires the use of numerical derivatives, which carries additional risks regarding numerical stability.

Note: technically, the configurations defined via GIMLI will work fine in both codes even if they are not in equilibrium. However, if the setup is not force-balanced, care has to be taken when interpreting the linear results.