24 Commits

Author SHA1 Message Date
107c0fcd7e set fdfd sources in a way that catches incorrect j.e calculation 2024-01-17 23:30:08 -08:00
c3b6fd94a6 should be e dot j.conj() 2024-01-17 22:20:59 -08:00
44465f1bc9 modernize type annotations 2023-05-22 10:53:13 -07:00
52df24ad98 Major typing updates
Split field types to differentiate between complex and purely-real

Fix lots of numpy-related stuff
2022-10-04 17:17:44 -07:00
faecc79179 typing and formatting updates 2022-10-04 14:32:40 -07:00
db23cc1d6f style and typing fixes in tests 2020-10-17 17:48:58 -07:00
bc428f5e8e add more type hints 2020-10-16 21:46:04 -07:00
d13a3796a9 style and type fixes (per mypy and flake8) 2020-10-16 19:16:13 -07:00
a956323b94 move stuff under fdmath 2019-11-27 22:59:52 -08:00
caa6f995c7 rename SimResult classes to TDResult and FDResult 2019-11-22 00:56:35 -08:00
63ad67800a further improve solve precision 2019-11-21 20:13:53 -08:00
0f07fd15da whitespace 2019-11-12 01:11:58 -08:00
41be4d2ab8 more cleanup 2019-11-12 01:11:13 -08:00
318f361f61 some cleanup 2019-11-12 01:10:58 -08:00
d4bb3e83e1 improve solution accuracy 2019-11-12 01:10:21 -08:00
b24b5d169f Add residual testing 2019-11-12 01:10:01 -08:00
2300a8b47e poynting equality needs bigger tolerance
not sure why yet
2019-11-04 20:49:04 -08:00
8794cbd7c2 e dot j should be masked as well 2019-11-04 20:48:14 -08:00
ca70d6e6c8 Sample all three S components... 2019-11-04 20:31:02 -08:00
68f857e83e Use 2-point sources to test fdfd poynting 2019-11-04 20:30:45 -08:00
0bb4105e7a e dot j should take into account dxes 2019-11-04 20:30:07 -08:00
9253200eaf Move j_steps and j_distribution out of conftest
- also move PRNG back into utils
2019-11-04 20:27:22 -08:00
60961db2d3 add atol to shut up solver 2019-10-27 16:17:15 -07:00
b0b295d1ac add test_fdfd 2019-10-27 16:13:25 -07:00