WENO

WENO

The focus of this group was to get WENO reconstructions working on 1d mapped grids. This was accomplished by:

1. Generating generic WENO routines (of various orders) that require smoothness indicator coefficients, optimal weights, and reconstruction coefficients to be supplied at run time and including them in SharpClaw.
2. Modifying PyClaw to detect when a nonuniform grid is being used, and subsequently calling PyWENO to compute the above coefficients and weights. These are attached to the PyClaw state object and passed down to the SharpClaw reconstructor if necessary.

The 1D homogeneous acoustics example was used as a simple test case. The example runs to completion and the solution appears to be correct when a nonuniform grid is used. Further testing should be done.

Unfortunately computing the smoothness indicators is quite time consuming (I am going to include a cached set of smoothness polynomials in PyWENO to speed this up).

Git repo

The 'hpc3-weno' branch (which is the default) of:

$ git clone [email protected]:memmett/pyclaw.git

contains PyClaw related code. A fresh version of PyWENO is also necessary.