pyro2

pyro2 copied to clipboard

Right now, we rely on the state machine interface to matplotlib. We should instead have the simulation class create and store a figure object and then work from that. Then...

zingale

For the `rt` problem, we default to HSE BCs, but these don't know about the 4th order reconstruction. We should use reflecting with the one-sided reconstruction in this case.

zingale

At reflecting boundaries, we want to use the one-sided reconstruction, like we do in Castro.

zingale

We should rename `simulation.cc_data` in `simulation_null.py` and all derived classes to just `simulation.data`. It is not cell-centered for the 4th order solvers, so it is a little confusing as written.

zingale

run codespell on notebooks

1

this will preprocess the notebooks and remove the output first

zingale

[WIP] implement one-sided stencils for 4th order

2

these are for symmetry boundaries.

zingale

The main `compressible` solver uses characteristic tracing, so the well-balanced scheme in `compressible_rk` is not complete. We should add the version from ppmpy (https://github.com/python-hydro/ppmpy)

zingale

zingale

this now allows stuff to flow out but not in also tweak the initial conditions for convection

zingale

This is likely because of the conversion of the initial conditions to averages right out of the gate. We need to do some flooring there.

zingale