-
Restructured initial conditions to have python scripts generate particle distributions instead of doing so inside the C++ code. This improves the code's flexibility.
-
Add option to output data in HDF5 format. Post-processing scripts only work with the original binary format, since yt only reads the binary format.
-
Add realtime output of scalar quantities to make basic analysis many times faster than with the post-processing scripts.
-
Include all of the basic post-processing scripts with Emu itself to avoid keeping multiple incompatible copies of them.
- Incorporated various feature additions used for Code Comparison for Fast Flavor Instability Simulations (https://doi.org/10.1103/PhysRevD.106.043011)
-
Using the new particle shape factor implementation, we now redistribute particles only at the end of each timestep instead of at every runge-kutta stage (#50).
-
Removed WarpX bilinear filter, since we don't need it to get stable evolution. Emu now does not include any WarpX code (#49).
-
Removed WarpX shape factors. Wrote new shape factor implementation in
ParticleInterpolatorso we now can evolve particles within ghost cells (#48). -
We now determine the shape factor order taking into account both the user-specified order and the dimensionality of the problem so we use order 0 in the directions where the domain contains only one cell (#47).
-
Fixed issue with flavor vector length so we can now run simulations with nonzero initial amounts of non-electron flavor neutrinos (#44).
- Emu has been through the LBNL licensing process for open source code. We release Emu under the BSD-3-Clause-LBNL license, the code is identical to v1.0.
- Version of Emu used for paper 1 Particle-in-cell Simulation of the Neutrino Fast Flavor Instability (https://arxiv.org/abs/2101.02745)