Documentation
The Llama documentation is currently a work-in-progress. For
details of the algorithms, see the Pollney et al. paper.
There is basic documentation for several of the Cactus thorns:
Coordinates |
Defines the patch systems |
[HTML] |
[PDF] |
Interpolate2 |
Performs interpatch interpolation |
[HTML] |
[PDF] |
GlobalDerivative |
Finite difference operators |
[HTML] |
[PDF] |
Slides
from the Einstein Toolkit Seminar on 04-02-2013 by Christian
Reisswig are also available, and provide some up-to-date details about
recent developments in Llama.
Compatibility
The following are already compatible with Llama:
- Any code which only accesses gridfunctions pointwise, and uses the coordinate gridfunctions x, y, z rather than assuming a uniform grid spacing. i.e. most initial data thorns will work unchanged (TwoPunctures).
- Any code which accesses gridfunctions only through the interpolator; the interpolator (via Carpet) is Llama-aware (e.g. AHFinderDirect, Multipole, PunctureTracker)
- Evolution and analysis codes which have been explicitly modified to work with Llama (McLachlan, WeylScal4, GRHydro, CTGamma, several Llama example codes)
- Any Kranc-generated code can be Llama-enabled by adding the option UseJacobian -> True to the CreateThorn function in the thorn script
- Output codes (CarpetIOASCII, CarpetIOHDF5 etc)
- The Carpet VisIt plugin
The following will need changes to work with Llama:
- Evolution codes performing non-pointwise operations (e.g. finite differencing)
- Anything that explicitly assumes that the local grid points are uniformly distributed in Cartesian coordinate space