Changelog

0.18.0 (2024-11-20)

Features

Scaling (#229) (faf792a)

Bug Fixes

BC no longer required to assemble the matrix of a local scheme (#231) (2080c00)
fill corners on uniform meshes (#234) (59d81b1)
fix block sizes for nested matrices (#233) (142be77)
fmt usage with version 11 (#226) (b3902b0)
The matrix size is set upon calling assembly.set_unknowns(...). (#227) (5d324de)

Performance Improvements

update_ghost_mr() performance improved on uniform meshes (#230) (2596ba7)

0.17.0 (2024-09-10)

Features

Boundary region with coordinates (#221) (5bcb38e)
export samurai CMake options (#224) (df0ded2)
Flux operators applicable by direction (#223) (91a7082)

0.16.0 (2024-06-18)

Bug Fixes

Fix compilation error on Get (#215) (d855db0)
Fix some minor issues in the flux-based schemes (#217) (c15b06d)

0.15.0 (2024-06-12)

Bug Fixes

create missing prediction cells outside the subdomain (#212) (fa693e0)
Fix flux-based schemes for MPI (#209) (768a5f4)
issue with get_cell function in LevelCellArray when dim > 2. (#211) (f5fe326)
keep_only_one_coarse_tag function (#210) (d90957a)

0.14.0 (2024-05-23)

Features

Jacobian matrix for the flux-based operators (#202) (95c2877)
OpenMP for independent Newton solvers (#201) (712a490)

Bug Fixes

add fmt:: prefix to function format_to (#204) (2b665ac)
fix row_shift and col_shift for monolithic assembly of an operator sum (#205) (f44995f)

0.13.0 (2024-05-06)

Features

Block matrix improvements (#198) (7b7c690)
Diffusion operator with different coefficient by field component (#194) (0a53b5e)

Bug Fixes

data structure xtensor by default if existing CMakeCache (#192) (bc90aef)
detail operator with samurai_check_nan compilation option (#190) (08cdb9d)
Fix compilation issue on non-linear cell-based scheme (#193) (df0ac4f)
MPI (#199) (d599908)
periodic (#188) (485377a)

Performance Improvements

change of FluxValue<cfg> structure (#186) (b1b8bf2)
Vectorization of linear homogeneous operators (#182) (153b2f2)

0.12.0 (2024-04-06)

Features

multi resolution adaptation (#171) (60d5892)
the direction provided in the boundary conditions (#183) (8bcb601)

Performance Improvements

Add MR ghosts only if level_min != level_max (#185) (57d5b73)
dynamic regions in extrapolation (#184) (16eef37)

0.11.0 (2024-03-20)

Features

Add precision (float-->double) in XDMF files (#179) (6b28811)
check for NaN in the computations (#176) (3954711)
High-order Dirichlet (#178) (3640062)
OpenMP on fluxes (#180) (57f2de7)

Bug Fixes

Fix advection_2d_user_bc (#177) (8d33411)

0.10.0 (2024-02-29)

Features

add dim in field_operator_base (#168) (d74b6ca)
additional initial value parameter to make_field (#162) (3536eac)
Boundary conditions with configurable stencil (#167) (e59c2a4)
collapsable flux value in non-linear flux-based schemes (#165) (bca0474)
const field in FV operator lambdas (#164) (7a78d2b)
local newton solvers (#153) (699468e)
Non-conservative fluxes (#161) (61d1cfd)
Parallelization using MPI (#28) (a411c76)
use xexpression for get_interval, get_index and get_cell (#170) (9391d2d)
WENO5 scheme (#166) (55fe777)

Bug Fixes

cell array instead of cell xtensor for the stencil of non-linear cell-based schemes (#156) (2487af3)
Fix flux-based schemes with unsymmetric stencil (#157) (697cca6)
Fix operators for time integration (#154) (7993525)
fixing (almost) all warnings triggered when compiling tests & demos (#163) (246ebcb)
right shift on interval (#169) (67f84e1)
simplifies, fixes and documents left and right shift of Interval (#172) (db8ce1c)

0.9.0 (2023-11-15)

Features

simplify bc and remove dispatch (#141) (7b06d90)

Bug Fixes

detail computation in MRAdapt (#148) (906c485)
non-linear schemes for scalar fields (#150) (bf80fa2)

0.8.0 (2023-10-19)

Features

Newton method (#135) (36b1244)
Portion for a slice (#136) (19e19ab)

0.7.0 (2023-10-09)

Features

add cell_in in functionBC (#121) (d6fcdc9)
Add direction on bc function (#131) (2d9ae52)
add Field_tuple (#116) (308abad)
Add precision (float-->double) in XDMF files (#119) (3187d1e)
add the possibility to change the field name (#111) (20a0e79)
heat N-D (#118) (c45c059)
linear heterogeneous fluxes (#126) (bfefd3d)
new flux-based schemes (#117) (1508453)
non-linear cell schemes (#130) (08a95af)
non-linear fluxes (#123) (25431d0)
remove field object from operators (#124) (8c11e00)
remove small entries in prediction_map (#115) (f0302a7)

Bug Fixes

BC function receives boundary coordinates (#122) (2b47dd7)
extract_mesh (#110) (3c2c05e)
Fix transfer() function for vectorial fields (#128) (5c58983)
heat + ghost elimination (#112) (5e43087)
index for upwind variable scheme to recover i/j symmetry (#120) (48d7e58)
missing coefficient for Neumann BC (#114) (b3de2cb)

0.6.0 (2023-07-09)

Features

field constructor from expression (#107) (0375ed9)

Bug Fixes

bc (#106) (f955a57)
coarsening when a parent intesects the boundary (#105) (4aa6411)

0.5.0 (2023-07-05)

Features

explicit schemes (#100) (c73cfd3)
Field iterator (#102) (4551eae)
own bc (#103) (7d92323)

0.4.0 (2023-07-03)

Features

add transfer function (#95) (d778fad)
create mesh from CAD object (#51) (8c53fd1)
scheme definition outed (#96) (0c62af4)

Bug Fixes

const field (#98) (8387396)
monolithic Stokes assembly (#94) (358fc37)

0.3.0 (2023-05-22)

Features

monolithic block matrices (#89) (6e318c2)

Bug Fixes

bc when ghost_width > 1 (#90) (611f1a1)
clang-tidy error fixes (#88) (bd131fa)
cpp version in target (#92) (126cf34)

0.2.0 (2023-04-22)

Features

convergence order computation (#83) (50d8419)
unification of the boundary conditions (#82) (647c314)

Bug Fixes

small fixes for vcpkg port (#85) (0172cb2)

0.1.1 (2023-04-18)

Bug Fixes

set a name for check-pr-title action (#80) (7e72642)