Inclusions example (2D acoustics on mapped grid).

[ketch]

This example corresponds (almost) to Figure 8.6 of the Calhoun-Helzel-LeVeque SIREV paper (2008). It uses the new solvers in clawpack/riemann#98.

This now has a test and is ready to merge.

[coveralls]

Coverage decreased (-3.89%) to 22.67% when pulling 4d6cccc on ketch:acoustics_mapped_2d into 995dd40 on clawpack:master.

[ketch]

@mandli This is ready to merge; could you please review it and clawpack/riemann#98?

[ketch]

Note that I used the .npz (compressed numpy array) file format for the regression data, drastically reducing the file size (versus text). I think this is the way to go for regression data, since I don't see myself ever opening the text file.

[coveralls]

Coverage increased (+0.01%) to 26.57% when pulling 084dc4f on ketch:acoustics_mapped_2d into 995dd40 on clawpack:master.

[ketch]

By the way, I think the compute_geometry function, which computes the face normals, edge length ratios, and cell area ratios should probably be merged into the Grid class. Having that code in Python seems fine as long as it is only called once (to set the aux array).

@weslowrie Has included corresponding code for 3D -- adapted from Grady's paper, I believe -- in clawpack/riemann#85.

[mandli]

Looks great! I think my only comment would be to fix the colormap's bounds so that it does not change while plotting.

[weslowrie]

@weslowrie Has included corresponding code for 3D -- adapted from Grady's paper, I believe -- in clawpack/riemann#85.

Yes @ketch that is the commit for 3D general geometries. I used Grady's paper, and LeVeque's FVMHP book for reference. For the 3D Euler Riemann solvers in general geometries, this Rohde paper was useful.

@article{rohde1999eigenvalues,
  title={EIGENVALUES AND EIGENVECTORS OF THE EULER EQUATIONS IN GENERAL GEOMETRIES},
  author={ROHDE, AXEL},
  journal={a a},
  volume={1},
  pages={1},
  year={1999}
}

[ketch]

@mandli Thanks for the suggestion -- I have done that and used a better colormap.

[mandli]

Looks great to me.