Commit Graph

717 Commits

Author SHA1 Message Date
Marek Nečada fa803573f0 Update Triton-troubleshooting.md
Former-commit-id: e092e45d4888a361dfb9c8f65cc68764d55411c3
2019-04-05 08:34:42 +03:00
Marek Nečada bf04cb32c0 Fix some earlier ewald32->ewald3 renames; disable some legacy code.
Former-commit-id: 6358a491f1965108a2747a9f48054a0022bcadf6
2019-04-03 16:41:10 +03:00
Marek Nečada 4212720560 Some docs to get qpms properly running on Triton. Needs an independent tester.
Former-commit-id: 62510002ceb858f2af0d4712324bb3749b0e904b
2019-04-02 17:46:06 +00:00
Marek Nečada c0b2343b89 Try whether Triton behaves better if the parallelized code uses no OpenBLAS.
Former-commit-id: ac1c17635e2d01d8a14129a90ffb43d3156510fa
2019-03-29 13:17:28 +02:00
Marek Nečada 652ab6f3ba Debug message types
Former-commit-id: 666a031026a363b15c01ca8cdc174fbf152dee8e
2019-03-27 09:25:47 +02:00
Marek Nečada eddb8558d8 Fix argument type.
Former-commit-id: 01b4413c9a09882b7da040bbb2eb6fb6dd0864dc
2019-03-26 15:23:17 +02:00
Marek Nečada 38bd16f94a Merge branch 'master' of necada.org:~/repo/qpms
Conflicts:
	misc/201903_finiterectlat_AaroBEC_fatMie.py

Former-commit-id: 6e9130313939be2eec80f4df9a703c8e9cb42456
2019-03-26 13:37:16 +02:00
Marek Nečada 9b50196ae6 AaroBEC scripts touch files indicating finished job.
Former-commit-id: 0786fba4eb57bcce086de0f123d53a80606e5bed
2019-03-26 13:35:30 +02:00
Marek Nečada 2ab4c55e13 Some control over the launched threads in scatsystem.
Former-commit-id: a67102dcb367b8fe577f9d8896c9a477571c7d6f
2019-03-26 11:33:22 +00:00
Marek Nečada f23edf0d71 Default number of threads in scatsystem.c set to 4, debug messages.
Former-commit-id: da244d95534eabe2be07327ba174467935518a12
2019-03-26 11:54:07 +02:00
Marek Nečada 5b49dfc8d0 Mie rect lattice script updates
Former-commit-id: 5ba256ee2f1489ce3bbbcfa9ee7da241bf2bff68
2019-03-25 20:31:40 +02:00
Marek Nečada 89df52061d Fixes in the new finite rectangular lattice / Mie particle script
Former-commit-id: 9e2b2f5987d7984d6adbb2ee364be1e7e4bf648f
2019-03-25 15:33:14 +02:00
Marek Nečada 4f4d340f0e Spherical particle version of the finite rectangular lattice script.
Former-commit-id: f23cb1a92cec73f38cd0d79c90346797e4c16b67
2019-03-24 23:58:32 +00:00
Marek Nečada f04ea1d600 Refractive index database into C incomplete draft.
Former-commit-id: fbfce8297f16400954fa3a1a1d24b980b67d0453
2019-03-24 20:19:55 +02:00
Marek Nečada f04fd9496e Additional functionality to Cython wrappers.
Former-commit-id: 69c34b8ff0fb2d59ee6e62414e9999f2717ecfa0
2019-03-24 19:30:33 +02:00
Marek Nečada 8683c69c0e Cython wrapper over qpms_permittivity_interpolator_t.
Former-commit-id: 38b362b52a8e8f5b6defca688b0589b25a2dfe22
2019-03-24 14:47:42 +00:00
Marek Nečada d7a64f5928 Reimplementation of qpms_permittivity_interpolator_t.
Former-commit-id: 740c89d441f3970bdd2aa4d64d8707c74e990b0c
2019-03-24 14:28:09 +00:00
Marek Nečada 115d1d07fe Material data interpolator + Mie T-matrix convenience functions.
Former-commit-id: 59c9a53bff38a46e357f091c084655192faddb51
2019-03-20 23:19:29 +02:00
Marek Nečada eafd1609a5 Using d1mach.f, i1mach.f copied from scipy/scipy/special/mach/.
The complex Bessel functions are finally working.


Former-commit-id: 6a1b896fc2cef938cb88bf8dbc6170f501ccf119
2019-03-20 21:02:28 +02:00
Marek Nečada 6ee4518bb1 Complex Bessel functions using amos.
Not working yet, probably because of unset machine constants in i2mach.f, d1mach.f.


Former-commit-id: 94a15b3bbbcb26b21533866c984076aeefcecb5e
2019-03-20 20:48:01 +02:00
Marek Nečada 959bb09526 More CMakeLists.txt additions.
Former-commit-id: a013013072fc854abcc7e5cc8366e1979ce4d3ba
2019-03-20 14:13:41 +02:00
Marek Nečada 522a6d9b85 Replace original netlib amos routines with those in scipy.
Note: apparently, one could also use amos from openspecfun
https://github.com/JuliaMath/openspecfun

Former-commit-id: 77d61fe6486803684201480410021fd173172cc0
2019-03-20 12:28:11 +02:00
Marek Nečada e6e7485ebb Futile attempts with cmake + fortran
Former-commit-id: da23f12faf79117152e28ef47185968dfddd5594
2019-03-20 09:28:47 +00:00
Marek Nečada 94259e766b Add AMOS routines from netlib
Former-commit-id: c6fe0e468937daa36fe5f6692af5d24d4df3f851
2019-03-19 14:42:47 +00:00
Marek Nečada 31bab5225f Initial CMakeLists.txt
Former-commit-id: a5ddf172f707e94cb7d823750ff0e2bb0524e432
2019-03-19 14:33:44 +00:00
Marek Nečada 4475cffeba C Mie T-matrices but demand yet Bessel functions of complex arguments.
Former-commit-id: 101201f5d72ab0f45eefc09c860b65b6841460bf
2019-03-18 19:54:29 +02:00
Marek Nečada 2af9b83e35 Move T-matrix -related C code to scatsystem.[ch]
Former-commit-id: e9a162c14fe5d91281d79d9af08014787cd7ed13
2019-03-18 15:04:21 +02:00
Marek Nečada 791a46b446 Fix M, N reordering in python T-matrix loader.
Former-commit-id: dd9709a5c82c19c33877e508a3336dce43c35b67
2019-03-18 09:35:10 +02:00
Marek Nečada 98da54946c Parallel mode problem matrix creation.
Former-commit-id: a81323f6ba53fc5e790e489314bca98f6ae1ca0a
2019-03-18 01:06:04 +00:00
Marek Nečada 86b87c955c Fix the alternative order modematrix implementation
Former-commit-id: 7daaf636ca803565e2a72b03e1ce184ece24e4c0
2019-03-17 23:32:51 +00:00
Marek Nečada 7db7123c91 Alternative implementation of irrep packed modeproblem matrix.
Former-commit-id: ed37e98423e03f96f6fb850820bfae451f505f21
2019-03-17 21:49:10 +00:00
Marek Nečada 96a6fdf7fa Scatsystem metadata to iterate the particles in the orbit-layout order.
Former-commit-id: 468f8c3ef64c891ae8f0ef415c6f30a42307d471
2019-03-17 16:46:22 +00:00
Marek Nečada 3e94b1baec The quaternions were apparently correct after all (except for the ordering of conjugation in vector rotation).
Former-commit-id: c233f1501a565ad3bbed5933324a0b821d2d1379
2019-03-15 21:14:59 +00:00
Marek Nečada 97f87202e1 Fix lowest SV index finding
Former-commit-id: ac6e4f3539dfa32888497600a8adf603d3587883
2019-03-14 01:34:04 +02:00
Marek Nečada ea0b6fb402 Rectangular lattice SVD computation script for the new paper.
Former-commit-id: 771b278b2eee810892934f379ceefc8554e24946
2019-03-13 23:44:12 +02:00
Marek Nečada 7ef8e764c7 Deterministic ordering of irreps in generated C structures
Former-commit-id: 052e894a00d56492af3a8481a355cfff7f84cfa5
2019-03-13 23:13:41 +02:00
Marek Nečada 51d38798c3 Some useful ScatteringSystem class functionality.
Former-commit-id: c7ae5dcd3f601b79781b685804dae5262fc2a93d
2019-03-13 22:35:35 +02:00
Marek Nečada 40caf44a24 Remove unused variables
Former-commit-id: d1bd6585b5a33d713969ecea3be9a9ffd5ec122c
2019-03-13 18:15:30 +02:00
Marek Nečada 412d520e13 "Direct" generation of irrep-packed mode problem matrices.
Former-commit-id: e0833046f2f6052e5e0523dce6047e89d23509ff
2019-03-13 18:13:59 +02:00
Marek Nečada a6dcd35b8d Revert setup.py to standard blas names and increment version.
Former-commit-id: e98c7ba026caefd85fd42288fb7e1987c96e5ad1
2019-03-13 16:23:23 +02:00
Marek Nečada 3f32b21730 Merge branch 'fixprojs'
Former-commit-id: 9e2357e6cc6844360b51a55a8af984dec82167a7
2019-03-13 16:21:45 +02:00
Marek Nečada 12da9f3b5a Various tests and macro conditions.
Former-commit-id: ed964c6284d98ed315a442be48b2dab3d7685222
2019-03-13 16:18:47 +02:00
Marek Nečada ddfd4954dc Fix complex conjugation; now all available point groups seem to work.
Former-commit-id: 764155cddfd30162b01c6552cea8718d55a26fd6
2019-03-13 15:59:05 +02:00
Marek Nečada 07aed6feb7 Fix offset increment in matrix packing.
Now it works for a z-dipole system.


Former-commit-id: 3e4aa9d4736ecc18e561d0ba5c2c19b0a027e37c
2019-03-13 13:24:44 +02:00
Marek Nečada d11db980f9 A slow but working reference implementation of matrix (un)packing.
Former-commit-id: 5f85c4983a8754ebae03980242b896a79954a73e
2019-03-13 12:28:40 +02:00
Marek Nečada cc61a64313 Dump parts of the full matrix.
Former-commit-id: 53b84dc031fab778b089226be8b26d89e6f6a4a0
2019-03-13 11:36:52 +02:00
Marek Nečada aee4d2db2a sss3 dump packed matrices
Former-commit-id: 5b2f8a4ebe9829d63546ae91e24c25ec5966fe72
2019-03-13 11:24:09 +02:00
Marek Nečada 6f219b2c84 Dump orbit projection matrices in sss3.c
Former-commit-id: 57814624995124945704e8a1bfc52273845633af
2019-03-13 08:57:23 +00:00
Marek Nečada 8cba539684 Tenhle případ rozbíjí rozborku sborku na diagonále.
(Přitom prohození první a třetí částice v zadání dává správný výsledek.)


Former-commit-id: 0da6fd465c44e803d15164a98dfd6896ff97d363
2019-03-12 23:03:26 +02:00
Marek Nečada 02ec8804d0 Fix quaternion conjugation order on vector rotation.
Now it does not crash on C4v and the projectors seem to be indeed
projectors, but the translation operator decomposition is still wrong.


Former-commit-id: 6273c5a76b4e424f77b12514df209b995999b24b
2019-03-12 22:22:02 +02:00