9f1c072704
DRIVER: Use only few last steps to estimate the remaining time.
...
This improved the remaining time estimation in case the number of leafs
changes significantly.
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-08 22:35:38 -03:00
4f73c95616
BOUNDARIES: Use the same buffer for boundary copy.
...
The copy of boundaries between blocks at the same level always uses the
same size of buffer, so instead of using two of them, one for sending
and another for receiving, use just one for sending and receiving. This
will use the appropriate subroutine from MPITOOLS
exchange_arrays_same().
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-04 10:22:13 -03:00
787e07d572
BLOCKS: Rename %block to %meta in block_info structure.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-04 09:31:44 -03:00
f3ad7924ea
MPITOOLS: Use interface for exchange_arrays().
...
Implement two subroutines, one for exchange of arrays of different
sizes, and the second for exchange of arrays of the same sizes.
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-03 22:44:30 -03:00
95617866b1
MPITOOLS: Use MPI_IN_PLACE in all MPI_Allreduce().
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-03 22:39:33 -03:00
3a056d4ddc
HOSTS: By default use IMPI with Intel compiler.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-03 13:01:33 -03:00
5bbd6b1242
MAKE: Add option -fcheck=all for DEBUG.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-03 12:41:19 -03:00
9ecbf83d5f
MESH: Make sure again the index nc is not out of the range.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-03 12:32:48 -03:00
192d417431
HOSTS: Add a few check flags to NVIDIA compiler options.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-03 11:33:57 -03:00
01d0549bf5
MESH: Make sure the index nc is no out of the range.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-03 10:32:23 -03:00
20cb799fba
MPITOOLS: Fix argument intent in exchange_arrays().
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-01 13:11:35 -03:00
e352c2cb0e
HOSTS: Add support for NVIDIA HPC compilers.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-11-01 11:33:50 -03:00
2098661d3c
EQUATIONS: Remove cruft.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-30 16:59:32 -03:00
8d5f356d8d
MPITOOLS: A real value and array are the same for NVIDIA compiler.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-30 16:53:35 -03:00
0ff346597d
HELPERS: fnum() is GNU Fortran specific.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-29 11:55:43 -03:00
f567a9ddf8
USER_PROBLEM: Pass the simulation time to user_time_statistics().
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-26 12:01:01 -03:00
b0401e96c6
PROBLEMS: Pass the run count to user problem initialization.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-26 11:59:11 -03:00
6187c6d76d
USER_PROBLEM: Rename user gravitational acceleration subroutine.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-26 09:09:13 -03:00
8c808fd70c
USER_PROBLEM: Rename subroutines boundary_user_() to user_boundary().
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-26 09:02:43 -03:00
22455676f4
USER_PROBLEM: Add user_time_statistics() to module USER_PROBLEM.
...
This file can we use to calculate and store the user defined time
statistics.
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-26 09:00:03 -03:00
94a621a3c5
USER_PROBLEM: Rename use problem subroutines.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-26 08:52:03 -03:00
6362fe13ad
EVOLUTION: Fix partial times and timesteps in 3S*+ methods.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-23 20:52:03 -03:00
8297482639
IO: Store 'cmax' so we can properly resume jobs.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-23 11:23:58 -03:00
44ea63bc3d
EVOLUTION: Implement optimized embedded methods by Ranocha et al.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-20 18:08:54 -03:00
27041d0feb
DRIVER, EVOLUTION, IO: Rewrite handling precise snapshots.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-19 14:03:57 -03:00
55b0d24b77
EVOLUTION: Rework slightly the way new time step is calculated.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-19 13:28:25 -03:00
c604467a5c
DRIVER: Use the initialize_time_step() instead of new_time_step().
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-19 13:15:04 -03:00
d501b19041
EVOLUTION: Implement estimation of the initial time step.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-19 13:11:41 -03:00
6ca36c93e3
DRIVER, EVOLUTION: Move dtnext to evolution module.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-18 11:33:14 -03:00
0c8d8fac9c
EVOLUTION: Rewrite SSPRK4(2)M integration method.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-16 19:03:25 -03:00
e7612f5cfd
PYTHON: It should be 'block x of y'.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-15 23:03:26 -03:00
418a580888
EVOLUTION: Print 'err/tol' not the error during the integration.
...
This is more meaninful, since it shows the ratio of the maximum error to
the tolerance atol + rtol * max(U).
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-15 15:29:42 -03:00
155b5d7575
EVOLUTION: Implement L2 and maximum norms for error calculation.
...
Alow the user to choose between these two norms.
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-15 13:51:25 -03:00
271af020df
EVOLUTION: Implement 3rd order embedded SSP3(2)4 with error control.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-15 12:15:14 -03:00
27b6daebca
PYTHON: Version 0.9.5
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 18:06:43 -03:00
0d4e955eb0
PYTHON: Reduce dimensions for 2D datasets in AmunH5.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 18:06:07 -03:00
9a7235b72d
PYTHON: Use None for fields in chunks with dblocks == 0.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 18:05:58 -03:00
51cc3ea83b
PYTHON: Determine toplev only for chunks with dblocks > 0.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 18:05:46 -03:00
d217d092ab
PYTHON: Version 0.9.4
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 13:31:05 -03:00
e13bd00a0e
PYTHON: Replace numpy.empty with numpy.zeros in AmunXML.__fill_chunks().
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 13:07:57 -03:00
07fc5c20b0
PYTHON: Create empty chunk fields for chunks with no dblocks.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 13:02:31 -03:00
5eec85e598
PYTHON: Version 0.9.3.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 10:06:16 -03:00
ec8ddac795
PYTHON: Add velocity and magnetic field components.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 10:05:18 -03:00
f4df543b2c
PYTHON: Import numpy without abbreviation.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 09:44:49 -03:00
b7496e0b3d
PYTHON: Make sure dataset() accepts maxlev > 9.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-06 09:37:15 -03:00
c2c9c37b45
PYTHON: Version 0.9.2
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-05 11:12:07 -03:00
6945b90b66
PYTHON: Export WriteVTK through the AmunPy interface.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-05 11:11:17 -03:00
5a8526ef82
PYTHON: Version 0.9.1.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-05 10:59:39 -03:00
1c4fa6035d
PYTHON: Rewrite vtkio.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-05 10:58:44 -03:00
e153cd417e
PYTHON: Rewrite interpolations.
...
Signed-off-by: Grzegorz Kowal <grzegorz@amuncode.org>
2021-10-05 10:19:18 -03:00