Skip to content

Latest commit

 

History

History
215 lines (127 loc) · 6.04 KB

CHANGES.md

File metadata and controls

215 lines (127 loc) · 6.04 KB

4.5.0

  • add protection against negative density/energy in the 4th order compressible method when converting from averages to centers (#309)

  • implement a "clean state" routine for the compressible solvers that can enforce a density floor (#329)

  • updated the citation information (#324) and the zenodo file (#325)

  • code cleaning (#323)

  • added python 3.13 support (#321)

  • CI doc build fixes (#322) and added the ability to skip executing notebooks when building docs (#319)

  • fixed issues with how derived variables are recognized when reading in an output file and add vorticity and machnumber to compressible derives (#294, #320)

  • expand the documentation on the compressible solvers (#317)

  • remove unused APIs from docs (#318) and fix autodoc issues (#316)

  • implement a sponge term in the compressible solvers (#313, #336)

  • add some asserts to the compressible solver to catch unphysical mass and energy (#312)

  • add the compressible convection problem (#293, #310, #315, #337, #338, #339)

  • add a multimode Rayleigh-Taylor problem to the compressible solvers (#307)

  • improve the documentation on working with data (#304, #305)

  • spherical compressible solver fixes (#295, #296, #298, #299, #300)

  • move the 4th order reconstruction into mesh (#302)

  • use a safer divide in compressible when density is zero (#297)

  • add the ability to have a problem-dependent external source and clean-up how the compressible sources are treated (#289, #290, #292)

4.4.0

  • the gresho problem (compressible) now takes Mach number as input instead of pressure (#287, #288)

  • doc updates to reflect how we run with the Pyro class (#266, #276, #284, #285) and other general docs cleaning (#265, #268, #270, #271)

  • pylint CI update (#283); python version bump in actions (#281)

  • a new HLLC solver with corrections for low Mach number was added (based on Minioshima & Miyoshi 2021) (#282)

  • documenting and clean-ups of problem setups for each of the solvers (adding docstrings that appear in the docs) (#272, #273, #274, #275, #277, #279, #280)

  • viscous_burgers was renamed burgers_viscous (#267, #278)

4.3.0

  • it is now possible to define a problem setup in a Jupyter notebook (#262, #264)

  • fix a bug in the artificial viscosity for spherical coords (#263)

  • I/O is disabled by default when running in Jupyter (#259)

  • the bc_demo test works again (#260, #261)

  • problem setups no longer check if the input grid is CellCenterData2d (#258)

  • the Pyro class interface was simplified to have command line parameter use the dict interface (#257)

  • problem setups no longer use a _*.defaults file, but instead specify their runtime parameters via a dict in the problem module (#255)

  • the compressible_sr solver was removed (#226)

  • gresho problem uses more steps by default now (#254)

  • the 4th order compressible solver only needs 4 ghost cells, not 5 (#248)

  • the compressible solver comparison docs were changed to an interactive Jupyter page (#243, #246, #249, #252)

  • some interfaces were cleaned-up to require keyword args (#247)

  • developers were added to the zenodo file (#242)

  • doc updates (#241)

4.2.0

  • moved docs to sphinx-book-theme (#229, #235, #240)

  • docs reorganization (#214, #221, #222, #234, #239) and new examples (#228, #236)

  • remove driver.splitting unused parameter (#238)

  • clean-ups of the Pyro class (#219, #232, #233) including disabling verbosity and vis by default when using it directly (#220, #231)

  • the advection_fv4 solver now properly averages the initial conditions from centers to cell-averages

  • each problem initial conditions file now specifies a DEFAULT_INPUTS (#225)

  • the gresho initial conditions were fixed to be closer to the source of the Miczek et al. paper (#218)

  • the RT initial conditions were tweaked to be more symmetric (#216)

  • the colorbar tick labels in plots were fixed (#212)

  • the compressible solver now supports 2D spherical geometry (r-theta) (#204, #210, #211)

  • the mesh now supports spherical geometry (r-theta) (#201, #217)

  • the compressible Riemann solvers were reorganized (#206)

  • CI fixes (#202, #215) and a codespell action (#199, #205)

  • python 3.12 was added to the CI (#208)

  • comment fixes to the compressible FV4 solver (#207)

4.1.0

  • Switched to pyproject.toml (#195)

  • pytest improvements allowing it to be run more easily (#194)

  • plotvar.py script improvements (#178)

  • a new viscous Burgers solver was added (#171)

  • a new viscous incompressible solver was added with a lid-drive cavity test problem (#138)

  • the incompressible solver was synced up with the Burgers solver (#168, #169)

  • convergence.py can now take any variable and multiplicative factor, as well as take 3 plotfiles to estimate convergence directly. (#165)

  • the multigrid solver output is now more compact (#161)

  • plot.py can fill ghostcells now (#156)

  • a new inviscid Burgers solver was added (#144)

  • a new convergence_error.py script for incompressible was added to make the convergence plot for that solver (#147)

  • regression tests can now be run in parallel (#145)

  • fixes for numpy > 1.20 (#137)

  • we can now Ctrl+C to abort when visualization is on (#131)

  • lots of pylint cleaning (#155, #152, #151, #143, #139)

4.0

This begins a new development campaign, with the source updated to conform to a standard python packaging format, allowing us to put it up on PyPI, and install and run from anywhere.

3.1

This is essentially the version from JOSS

3.0

This was the transition to Numba

2.2

  • added shallow water solver

  • added particles support

2.1

  • documentation switched to Sphinx

  • extensive flake8 clean-ups

  • travis-ci for unit test (with 79% coverage)

  • SDC compressible method added

2.0

This is the state of the project toward the end of 2017. 4th order finite-volume methods were just introduced.