A Bundle Tree-type implement of Adaptive Mesh Refinement (AMR) for General Relativity Magneto-hydrodynamics (GRMHD) simulations.
By analysis of traditional AMR frameworks, we find that what we're actually doing is not just adaptively refining meshes. Since we usually define our data on a customized mesh, we can regard this structure as a fibre bundle. As for adaptive refinement process, we can realize it by submitting refined grids into next or higher level (corresponding to AMR level). In this way, we obtain a tree-like bundles and name it naturally as BundleTree.
In this file, we pre-defined three useful classes for establishing a BundleTree for a specific problem.
To define grids or meshes, one can initialize a Chart object. To define corresponding data, one can use given data and newly created Chart object to initialize a Bundle object. In order to build a tree, a class named TreeNode is provided. You can refer to our test file test.py for more details.