-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathchangelog.txt
501 lines (315 loc) · 15.7 KB
/
changelog.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
release 0.2.1 (13 January 2012)
===============================
Bugfixes/Improvements
* Patch from [email protected], allowing from-source
installation in /opt
* Fix for mpi regression test
* Minor improvements for documentation (dispersion relation
computation and anisotropy examples in manual)
* Correction for instructions to run tests in nmag-core install.
* Adding python-pyvtk to list of required packages in installation
instructions.
release 0.2.0 (13 Dec 2011)
===========================
New Features
* Improved handling of command line arguments.
The parsing of command line arguments now happens at the beginning, before
the nmag module is imported. Options can be passed to PETSc using the
--petsc-opts flag (try `nsim -h --petsc-opts -h' to get a list). It is then
possible to use the logging facilities offered by the library. The
interactive command `nsim_i' has been removed. Nsim detects when it is
being used interactively and uses IPython when possible.
* Added a new command Nmagprobe for postprocessing the data. Can probe the
magnetisation in a cubic lattice in space and time. Can convert the data
into one or more OOMMF OVF files (both versions 1.0 and 2.0 are supported)
and these files can be reloaded with OOMMF, for example. Can Fourier
transform the data (in space, time or both) to obtain dispersion curves or
the modes of the system.
* Partial PBC support for HLib. HLib can now be used together with PBC.
Downside: the size of the HLib matrix increases with the number of copies.
For a sufficiently large number of copies (typically 1000 or more), using
PBC+HLib may not give any benefits with respect to PBC alone.
* Major restructuring of the installation procedure. In addition to
the installation from a large tar ball as in the previous release,
it is now easier to install Nmag using the system's Python and other
libraries (assuming the versions are compatible). This has been
tested for Ubuntu 11.10 and 11.04.
Bugfixes/Improvements
* Adjusted flags in compilation Makefile.
PETSc compiled --with-debugging=no (~10 % speed improvement). Stopping
when HLib compilation fails.
* Optimisation of mesh module: topology information generated only on demand,
mesh datastructures re-engineered to occupy less memory. As a consequence
a medium-large simulation requires less memory (reduction of 10-20 % of the
non-BEM part) and is faster in the initialisation phase (about the same
amount).
* Added a bug report script to collect compilation logs into a single file.
That should considerably simplify the submission of bug reports.
* Updated packages to latest stable release in the compile from source.
We are now using: Mpich2 1.2.1p1, Petsc 3.1-p5, Ocaml 3.12.0, Findlib 1.2.1,
Gsl 1.14, Parmetis 3.1.1, NumPy 1.5
* Added support for Numpy array to pass data to the core simulation engine
(makes it faster to set and retrieve fields)
* Fixed bug when reloading data from an MPI simulation with different number
of node with respect to the one which saved the data (data is now reordered
when loading it, if necessary)
* Reworked build system. All Nmag OCaml sources are now in one single directory
which makes it easy to configure and build the nsim executable.
* Native compilation on Ubuntu 11.10 should now be possibe
and relatively straightforward.
* Rewrote the test system and fixed bugs.
* Ported the documentation to Sphinx.
Release 0.1 beta (6481) 22 December 2009
========================================
New Features
* Support of Hierarchical Matrix compression for Boundary element matrix. See
example in documentation for details, and `Knittel et al 105, 07D542 (2009)
<http://link.aip.org/link/?JAPIAU/105/07D542/1>`__, `postprint pdf
<http://www.soton.ac.uk/~fangohr/publications/postprint/Knittel_etal_JAP_105_07D542_2009.pdf>`__).
Bugfixes/Improvements
* Minor improvements manual (Example 1 and 2, nmagpp, mayavi, 6170, updated MPI section)
* Updated plot.dat and plots for Example Hysteresis loop for thin disk to fix
documentation bug (thanks to Weiwei Wang for spotting this, 6326).
* The time stopping criteria are now more faithfully respected. A new optional
argument, 'exact_tstop', has been added to 'Simulation.advance_time' and to
'Simulation.set_params' to give more freedom to the time integrator, in case
performance is a concern and exact time-stopping is not really needed.
* Updated packages to their latest stable release in the compile from source.
We are now using: Sundials 2.3.0, HDF5 1.6.10, Python 2.6.4, pytables 2.1.2,
NumPy 1.3.0, IPython 0.10.
* The regression tests can now be called with three different targets (6430):
make check : runs all fast tests (should complete in less than 5
minutes). This is a good test to run after compiling nmag (from source).
make checkmpi : runs mpi-related checks
make checkslow : runs tests that take longer (order of hours)
make checkall : run all tests together
(Previously, the mpi tests were part of target 'check'.)
----------------------------------
Interface changes;
* We stop providing a Debian package for this release. The from source
compile works out of the box on Deian systems.
Release 0.1 beta (6163) 30 March 2009
=====================================
Bugfixes:
* Fixed bug that resulted in simulation hanging when running in MPI
mode (6039)
* Solved error in regression test (test_gmsh1_0_import.py). The error
was in the test code, not in the conversion code. (5987)
New Features:
* New functionality: Zhang-Li spin torque coupling term for uniform
current densities added and documented. (6133)
Improvements:
* Reduced memory footprint (by approx 7% for the problem in manual/example2)
(6098)
* Documented ``load_m_from_h5file`` (6041) and ``save_restart_file`` (6096)
* Document LLG equation of motion (new subsection in section
Background), and llg_gamma_G parameter in MagMaterials (6014)
* Added more ``save=`` usage examples for hysteresis() and relax()
functions (6096) in documentation of hysteresis command.
* Added MPI-regression tests (6031)
* tidied up file tree (removing out-dated examples, etc) (6059)
* Added mpi debug function nmag.mpi_hello() (6068)
* Allow to switch on debug messages from slaves (usually switched off)
using --slavelog
* Adding release date to output of nsimversion (6153)
Interface change:
* Configuration files (RUNID_*.conf) are now only written when the
--dumpconf switch is provided (6093)
Release 0.1 beta (5962) 20 November 2008
========================================
Bugfixes/Improvements/New Features
----------------------------------
* New feature: Can read finite element meshes produced with Gmsh using
the new ``--gmsh`` switch for the ``nmeshimport`` tool. (5924)
* Added more regression tests
* Fixed bug that stopped simulations where two magnetic domains touch
each other (5886)
* Fixed bug that didn't allow to have different (potentially
unrelated) simulation objects in the same nmag-program (5956)
* Fixed bug that (sometimes) stopped the --restart mechanism
to continue interrupted hysteresis loop calculations (5907)
* Fixed bug that prevented ``sim.get_subfield_positions`` and
``sim.get_subfield_sites()`` from working. (5933)
Release 0.1 beta (5878) 13 October 2009
=======================================
Bugfixes/Improvements/New Features
----------------------------------
* bugfix: reinitialisation used to reset the stage_time and stage_step
counters. This is now fixed and is done explicitly in the hysteresis method.
This functionality is needed to study systems where the applied field
changes in time (5848).
* bugfix: allow from source compilation on (K)Ubuntu 8.04. The bug is
with the Petsc library 2.3.2. We have upgraded the Petsc library to
2.3.3 in the compile from source installation, which resolves the
problem (5851).
* bugfix: reading Gambit meshes works without warnings (5852).
* bugfix: A missing ``.petsc`` file in the user's home directory
resulted in an error message when nsim was started. This did create
the missing file, so all is fine the next time nsim is
called. Revision (5853) fixes this.
* Improvement: Split hysteresis example in tutorial into two
examples. The first can be used to play around with thecommand andto
quickly complete a simulation. The second allows to compare the
resulting hysteresis loop with an analytical prediction (5876).
* minor improvements in makefile for installation from source (5859).
Release 0.1 beta (5787) 2 September 2008
========================================
Bugfixes/Improvements/New Features
----------------------------------
* bugfix: disabled diagnostic debug code that caused segfault for large meshes
(over 130,000 volume elements)
* bugfix: including documentation for
Simulation.get_subfield_average() in manual
* nmeshipmort has slightly improved error reporting when attempting to
import an unknown mesh format
Release 0.1 beta (5787) 8 August 2008
=====================================
Bugfixes/Improvements/New Features
----------------------------------
* We have a new distribution mode using VMWare's vmplayer. This is
particularly attractive to MS Windows users. This is at a very early
stage and and we invite feedback from early adaptors.
* Updated the Knoppix CD to the most recent version, and to include the
Netgen mesh generator.
* ncol can now process OOMMF's odt files (use --odt switch)
* Bugfix: the previous debian package could only execute on CPUs with sse2 extension.
* Bugfix: fixed embarrassing bug that appeared with the last release
(printing of applied field brakes when first component is zero).
Interface change
----------------
* removed support for thermal fluctuations. There appeared to be no
up-take of this feature, and maintaining this (while developing the
code further, parallelising it and making faster and leaner) was
unnecessarily binding resources.
Release 0.1 beta (5734) 1 August 2008
=====================================
Bug fixes/Improvements/New Features
-----------------------------------
* Reduced memory requirements by approximately 20% (due to better
compression of sparse matrices)
* Updated code working with hdf5 files to be compatible with pytables
version 1 and 2 (no immediate effect on user if using compile from
source, but allows to link to existing pytables2). By default we
are still using pytables1 (but expect to switch at some point in the
future, which should be transparent to users).
* Change from numpy 1.0.1 to 1.1.0. in compile from source distribution.
* More detailed progress reporting in nmagpp (useful if data files
become large, i.e. a GB and more).
* More detailed progress report in setup phase.
* More detailed progress reporting in hysterises/relax commands.
* Made compile from source work for Mac OS X 10.5 (Leopard)
* Introduced unit and regression tests.
Release 0.1 beta (5531) 2 June 2008
===================================
Bug fixes/Improvements/New Features
-----------------------------------
* The current release, distribution mode, svn release number and
location of the nmag-python libraries are provided (to stdout and
logfile) when a simulation object is created.
* Re-designed and implemented calculation of the Jacobi matrix: The
new approach
- slightly more computation time is spent during the main simulation
whenever the Jacobian needs to be re-computed
- the setup phase takes (notably) less time.
- simulations with the new approach need about 30% less memory
* Matrix setup is now parallised.
* Documented the use of MPI in the manual to carry out parallel runs.
* Bugfix: Code now working with MPICH2.
* Bugfix: The compilation from source works (again) for Mac OS X.
* Switched to pvode (the parallel solver for time integration) by default.
Updated timestepper tolerances section accordingly.
* Feature enhancement: can save spatially resolved fields using the
save_data command in any order (previsouly, it was possible to stick
to the same set of saved fields for every saved timestep).
* Update of manual in several places (including recommendation how to
cite nmag in FAQ).
* A number of (smaller) bugfixes.
Interface changes
-----------------
* The use of ``row`` for the analyses of nmag's ``_dat.h5`` data files
has been superseeded with the use of ``id``s (see explanation in
manual).
This has the following side effect: the nmagpp command labels vtk
file using the mask NAME-00XX.vtk where XX is an increasing
number. This used to be the 'row' number for saved fields. Because
it is now possible to save any fields in any order, the value of XX
is now replaced with the unique id of a saved configuration. This
also allows easier identification of a saved configuration with an
entry in the ndt file (via the id).
* nmagpp used to have a switch ``--field`` which allowed to choose
*one* field that should be dumped (``--dump``) or converted to vtk
files (``--vtk``). The functionality has now been extended so that
*a list of fields* can be given, such as ``nmagpp --fields
m,phi,H_demag --vtk myvtkfiles.vtk
mysimulation_dat.h5``. Consequently, the switch has been renamed
from ``--field`` to ``--fields``.
Release 0.1 beta (5259) 30 January 2008
=======================================
Bug fixes/Improvements
----------------------
* changed for parallel execution to use of pvode (rather than cvode
before). Results in performance increase for MPI runs.
Interface changes
-----------------
* None
New features
------------
None
Release 0.1 beta (5259) 30 January 2008
=======================================
Bug fixes/Improvements
----------------------
* ticket 126: Calculation of exchange energy, demag energy, Zeeman
energy and total energy had wrong sign; now working correctly.
* increased speed of setup phase slightly.
Interface changes
-----------------
* Have changed version of mpich in tarball distribution. Used to be
mpich2.0 but is mpich1.0 from now on (to be consistent with debian
packages). Users who have older tarballs and want to use MPI
execution, need to obtain a new tarball to update (updating the
nsim-core is not sufficient).
New features
------------
None
Release 0.1 beta (4947) 20 December 2007
========================================
Bug fixes/Improvements
----------------------
* ticket 101: Setting anisotropy=nmag.uniaxial_anisotropy(K1=0) in MagMaterial
produces an error message;
* ticket 103: When starting from a zero torque configuration the simulation
proceeds up to the maximum time (which is 1 s) and here it stops:
now an appropriate error message is displayed.
* ticket 97: Unrecognized architecture causes compilation failure in the
bem3d module: now the architecture flag will be empty if the architecture
detection fails.
* changeset:4919 reduced setup time by approx 1/3 through faster setup
of operators (apart from H_exch and Jabobi). Further speed up reduces setup time
to half the original value.
* new section on sundials tolerances in manual guided tour.
Interface changes
-----------------
None.
New features
------------
* Writing sundials performance data in log file (_cvode.log)
* added setting of CVODE tolerances as function of stopping_dm_dt
(needs to be tested and documented), currently switched off by
default.
Release 0.1 beta (4785) 27 November 2007
========================================
Bug fixes/Improvements
----------------------
* improved renderening of bitmaps (figures and equations) in html
documentation (4754)
* install from tar ball was broken (pkgs in wrong directory)
Interface changes
-----------------
New features
------------
* added mini tutorial on micromagnetic simulations to manual
* new tool nmeshsort
Release 0.1 beta (4625) 2 November 2007
=======================================
First release