Commit Graph

62 Commits

Author SHA1 Message Date
4e0bd8b3c6 switch to meanas 2020-10-16 19:32:22 -07:00
a4dd031666 ongoing 2019-08-30 22:13:26 -07:00
0c7b2fd3a2 add test_simulation 2019-07-30 00:33:35 -07:00
23d65d6ebb Add poynting halves
maybe will remove this later... dunno if it's actually useful
2019-07-30 00:33:07 -07:00
cf0db63a3f Revert "Add restrict keyword to pointers (not sharing the same memory for multiple fields)"
This reverts commit 60b70bb332.

It appears to have minimal performance impact, and fails to compile on
nvidia cards.
2019-07-17 00:53:55 -07:00
89d99187b6 Remove unused code 2019-07-17 00:52:31 -07:00
1627d0da05 Consolidate poynting code into update_h 2019-07-15 00:11:09 -07:00
fe2f4f4510 style fixes 2019-07-15 00:06:43 -07:00
314e36a3cc fix for 1-sized grids 2019-07-15 00:06:43 -07:00
a37df3a74f expand gitignores 2019-07-15 00:06:43 -07:00
jan
582dafbc2f Update example with bloch fields 2018-11-30 01:04:00 -08:00
jan
385e2859a9 Update parameter descriptions 2018-11-30 01:03:20 -08:00
jan
0f0f16e184 Minor formatting changes 2018-11-30 01:03:10 -08:00
a76e741d32 Minor spacing changes 2018-11-30 00:37:16 -08:00
60b70bb332 Add restrict keyword to pointers (not sharing the same memory for multiple fields) 2018-11-30 00:37:07 -08:00
9cf50b1d88 Enable nonzero pml alpha 2018-11-30 00:36:32 -08:00
65cc118392 Remove debug code 2018-11-30 00:36:22 -08:00
f5a5a0ad04 Enable nonuniform grids (minimally tested) 2018-11-30 00:36:11 -08:00
5ffe547640 Typo fixes and minor comment updates 2018-11-30 00:13:27 -08:00
3ac20f6271 Add bloch boundaries (untested) 2018-11-29 02:01:48 -08:00
d8dc024626 Implementation of "source" fields (J) 2018-11-29 02:01:28 -08:00
cb471df182 Implement proper kappa for PML 2018-11-29 02:00:30 -08:00
f00c8b4a3e Add _create_context(), _create_operation(), and _create_pmls(), and generalize initial field value args 2018-11-29 01:59:05 -08:00
2b1d906665 Add _create_field() and _create_eps() 2018-11-29 01:38:28 -08:00
1e874cb0c0 Fix sign on H component of PML 2018-11-29 01:36:13 -08:00
ea5e298023 Explicitly cast to int 2018-11-29 01:35:20 -08:00
1de6fb0e39 Use readme as long_description 2018-09-16 20:12:52 -07:00
d0b303523e Move version string into module 2018-09-16 20:12:41 -07:00
c8298d916f Use python3 for setup.py 2018-09-16 20:12:26 -07:00
aab57412a5 move code to new location 2018-01-15 22:36:40 -08:00
jan
a67009d1c8 fix declaration 2017-10-06 14:20:10 -07:00
jan
6a03977a69 fix pml param names 2017-10-06 14:12:51 -07:00
jan
9772f47a34 fix typo 2017-10-06 14:01:22 -07:00
jan
6bd756c3d3 add setup.py 2017-10-06 13:50:57 -07:00
jan
d02cd18403 improve pml specification 2017-10-06 13:49:46 -07:00
jan
c137da15b9 Merge branch 'master' of mpxd.net:jan/opencl_fdtd 2017-08-24 11:28:14 -07:00
jan
a85f547749 doc updates 2017-08-24 11:28:03 -07:00
jan
0d91f0d43e rename lib 2017-08-14 15:41:20 -07:00
97d9901e4b Use logging package for output 2017-08-12 19:20:29 -07:00
c1750e9fde Update readme 2017-03-29 01:09:21 -07:00
4633ababa5 Merge branch 'array_pml' 2017-03-28 22:08:11 -07:00
jan
d34c478f1d Rewrite, with the following features:
- Move to jinja2 templates for the opencl code
- Combine PML code into the E, H updates for speed
- Add Poynting vector calculation code, including precalculation during H update
- Use arrays for PML parameters (p0, p1)
- Switch to linearized, C-ordered fields (~50% performance boost??)

- Added jinja2 and fdfd_tools dependencies
2017-03-28 21:53:51 -07:00
jan
6c3313d7c9 call overhead still way too big 2016-09-01 14:45:32 -07:00
jan
a6e601b648 Clean up comments 2016-09-01 14:45:32 -07:00
jan
cd72219d0b Revert "lookup tables"
(accidental commit)

This reverts commit 7bf7a3d80c.
2016-08-05 18:40:27 -07:00
jan
7bf7a3d80c lookup tables 2016-08-05 18:39:58 -07:00
jan
481613fe5d Stop calling type_to_C(self.arg_type) all the time 2016-06-21 18:27:11 -07:00
jan
f4b2df2b2c Make some variables lower-case 2016-06-21 18:26:42 -07:00
jan
73aaa720b8 Remove extra spaces 2016-06-21 18:26:25 -07:00
jan
2018fa819e Cleaner representation of the field source 2016-06-21 18:26:16 -07:00