]>
https://gitweb.dealii.org/ - dealii-svn.git/log
guido [Tue, 17 Sep 2002 22:12:05 +0000 (22:12 +0000)]
new target for sparse_matrices in optimized mode
git-svn-id: https://svn.dealii.org/trunk@6448
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Tue, 17 Sep 2002 21:43:12 +0000 (21:43 +0000)]
new function BlockVector::collect_sizes
git-svn-id: https://svn.dealii.org/trunk@6447
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 20:05:12 +0000 (20:05 +0000)]
Doc update.
git-svn-id: https://svn.dealii.org/trunk@6446
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 19:26:43 +0000 (19:26 +0000)]
Add new data.
git-svn-id: https://svn.dealii.org/trunk@6445
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 19:26:13 +0000 (19:26 +0000)]
Add comments, and cosmetic corrections.
git-svn-id: https://svn.dealii.org/trunk@6444
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 19:12:56 +0000 (19:12 +0000)]
Join two loops, which is actually necessary, since we need some
information from the start of the loop in the second loop.
The mail that explains this is (sorry, only german):
---------------------------------
der shapes-Test geht bei mir fehl, und der Grund ist dass in einem
FESystem die _Werte_ nicht von den Basiselementen kopiert werden. Jetzt
habe ich diesen Teil gestern geaendert, so dass das wohl an mir liegt,
aber ich weiss nicht, warum es vorher funktioniert hat.
Also, wie haben ein Q2xQ3 FESystem. Fuer beide muss update_values nicht
in update_each gesetzt werden, weil sich die Funktionswerte ja nicht
aendern. Aber: irgendwann muss es halt mal ins FESystem kopiert werden.
Wo/wann passiert das?
Wenn ich naemlich das erste Mal in FESystem::compute_fill lande (von
FESystem::reinit auf der ersten Zelle), dann habe ich hier (das ist der
alte Code, der neue ist etwas anders, aber im wesentlichen der selbe):
for (unsigned int base_no=0, comp=0; base_no<n_base_elements(); ++base_no)
{
const FiniteElement<dim> &base_fe=base_element(base_no);
typename FiniteElementBase<dim>::InternalDataBase &
base_fe_data = fe_data.get_fe_data(base_no);
const UpdateFlags base_flags(dim_1==dim ?
base_fe_data.current_update_flags() :
base_fe_data.update_flags);
[da FEValues ist hier dim_1==dim]
[... und weiter unten beim kopieren:]
if (base_flags & update_values)
data.shape_values(system_index, point)=
base_data.shape_values(k,point);
Hier enthaelt base_flags kein update_values mehr, so dass die Werte nicht
kopiert werden.
Da das die einzige Stelle in fe_system ist, wo data.shape_values gesetzt
werden, macht mich das ganze ein bisschen stutzig: da kann doch nur was
kopiert werden, wenn die base_flags fuer diese Zelle update_values
enthalten. Also, nachgeschaut: base_fe_data.update_once hat
update_values, aber base_fe_data.first_cell==false...
Ah, jetzt sehe ich auch was passiert: vorher hatten wir nur eine einzige
Schleife ueber die Basiselement, und in der haben wir die Flags geholt,
das Element aufgefordert sich zu initialisieren, und gleich kopiert.
Jetzt haben wir zwei Schleifen: einmal initialisieren, dann kopieren. Ich
weiss damit auch, wie ich das Problem loesen kann (allerdings erst nach
dem Mittagessen), aber deshalb mal die Frage: ob wir auf dem ersten
Element sitzen, ist eigentlich nicht damit abgehakt, dass sich das
Basiselement initialisiert hat. Trotzdem machen alle Elemente das am Ende
des ersten Aufrufs von fill_fe_*_values. Ist das der richtige Platz?
Und: ganz kurios macht das das FESystem, naemlich irgendwo in der Mitte
der compute_fill-Funktion (die von den fill_fe_*_values Funktionen
aufgerufen wird), und zwar *im Innern eines Blocks, der in
if (flags & (update_values | update_gradients))
steht*!! Das will mir eigentlich nicht ganz richtig scheinen...
Ich werde nachher mal einen vorlaeufigen Fix einchecken, damit es
wenigstens wieder geht, aber wuerde trotzdem gerne Eure Meinung dazu
hoeren!
-------------------------------------
...and...
-------------------------------------
Zu ersterer: die update_flags werden auf eine absolut undurchschaubare Art
und Weise gesetzt, geaendert, von irgendwelchen Funktionen zurueckgegeben,
etc. Ich habe bis heute nicht verstanden, welche Funktion ich denn wann
aufrufen muesste, und dass Flags wie first_cell waehrend einer laufenden
(aeusseren) Funktion beeinflussen, was man von current_flags
zurueckbekommt, passt da sehr gut ins Bild.
Ich meine, mal ehrlich: ist es logisch, dass wenn ich current_update_flags
aufrufe, dann fill_fe_values, dann noch mal current_update_flags, dass das
Ergebnis dann ein anderes ist? fill_fe_values ist eigentlich ein typisches
Beispiel einer Funktion, die _Daten_, nicht den _Zustand_ eines Objektes
aendert/aendern sollte! Ich glaube, die Funktion hat bisher eigentlich nur
aus Zufall funktioniert, weil eine lokale Variable oben in einem Block
deklariert und definiert wurde, waehrend sie erst unten benutzt wurde.
Haette man die Deklaration dahin getan, wo die Variable das erste Mal
benutzt wird, dann haette das schon nicht mehr funktioniert.....
--------------------------------
git-svn-id: https://svn.dealii.org/trunk@6443
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 16:44:26 +0000 (16:44 +0000)]
Change comment.
git-svn-id: https://svn.dealii.org/trunk@6442
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 16:05:55 +0000 (16:05 +0000)]
Add an assertion that promptly triggers, and remove the underlying bug.
git-svn-id: https://svn.dealii.org/trunk@6441
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 15:48:11 +0000 (15:48 +0000)]
Reindent some parts.
git-svn-id: https://svn.dealii.org/trunk@6440
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 17 Sep 2002 14:46:53 +0000 (14:46 +0000)]
Add TODO note.
git-svn-id: https://svn.dealii.org/trunk@6439
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 22:12:01 +0000 (22:12 +0000)]
Get rid of component_to_system_table, and fix a couple of other problems with non-primitive elements as arguments to FESystem.
git-svn-id: https://svn.dealii.org/trunk@6438
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 22:06:19 +0000 (22:06 +0000)]
Things about restriction_is_additive.
git-svn-id: https://svn.dealii.org/trunk@6437
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 22:00:11 +0000 (22:00 +0000)]
Implement some things with gradients in 3d for the Nedelec element.
git-svn-id: https://svn.dealii.org/trunk@6436
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 20:08:10 +0000 (20:08 +0000)]
Don't set up component-to-system-tables when they are of no use,
i.e. for the non-primitive case.
git-svn-id: https://svn.dealii.org/trunk@6435
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 18:10:12 +0000 (18:10 +0000)]
Clarify comment.
git-svn-id: https://svn.dealii.org/trunk@6434
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 17:55:09 +0000 (17:55 +0000)]
Add this test, complementing non_primitive_1, but this time with Nedelec-elements. It is presently failing due to a known bug in FESystem, but this way it is in CVS, and I can care about the bug next.
git-svn-id: https://svn.dealii.org/trunk@6433
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 17:29:39 +0000 (17:29 +0000)]
Add some things.
git-svn-id: https://svn.dealii.org/trunk@6432
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 17:17:55 +0000 (17:17 +0000)]
Nedelec elements now work, but not systems of Nedelec elements.
git-svn-id: https://svn.dealii.org/trunk@6431
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 17:09:40 +0000 (17:09 +0000)]
Add a comment, and fix the 3d part.
git-svn-id: https://svn.dealii.org/trunk@6430
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 13:33:26 +0000 (13:33 +0000)]
Change the meaning of the restriction_is_additive flag: previously it was indexed on the vector component. With non-primitive elements, this does no longer work, so index it on the shape function number instead.
git-svn-id: https://svn.dealii.org/trunk@6429
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 13:29:37 +0000 (13:29 +0000)]
Change the meaning of the restriction_is_additive flag: previously it was indexed on the vector component. With non-primitive elements, this does no longer work, so index it on the shape function number.
git-svn-id: https://svn.dealii.org/trunk@6428
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 13:28:47 +0000 (13:28 +0000)]
Fix some of the last things.
git-svn-id: https://svn.dealii.org/trunk@6427
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 16 Sep 2002 13:27:24 +0000 (13:27 +0000)]
New test.
git-svn-id: https://svn.dealii.org/trunk@6426
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 22:42:38 +0000 (22:42 +0000)]
Add a test that checks the prolongation of Nedelec elements. The results are hand-checked, so should be correct.
git-svn-id: https://svn.dealii.org/trunk@6425
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 22:27:41 +0000 (22:27 +0000)]
Fix embedding matrices for 2d.
git-svn-id: https://svn.dealii.org/trunk@6424
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 22:16:07 +0000 (22:16 +0000)]
Add comment.
git-svn-id: https://svn.dealii.org/trunk@6423
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 22:15:05 +0000 (22:15 +0000)]
Remove unnecessary variable.
git-svn-id: https://svn.dealii.org/trunk@6422
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 21:59:53 +0000 (21:59 +0000)]
Constraint matrices are now correct. The nedelec_2 test checks them,
and I tested their values by hand computation, so they should be ok.
git-svn-id: https://svn.dealii.org/trunk@6421
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 21:58:51 +0000 (21:58 +0000)]
Add a second nedelec test, checking interface constraints for these strange elements. They are hand-checked, so should be correct.
git-svn-id: https://svn.dealii.org/trunk@6420
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 21:02:20 +0000 (21:02 +0000)]
Add comment.
git-svn-id: https://svn.dealii.org/trunk@6419
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 13 Sep 2002 17:01:29 +0000 (17:01 +0000)]
deal_II_numbers.
git-svn-id: https://svn.dealii.org/trunk@6418
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:46:44 +0000 (22:46 +0000)]
Unify the places where we use pi.
git-svn-id: https://svn.dealii.org/trunk@6417
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:44:15 +0000 (22:44 +0000)]
Fix a comment.
git-svn-id: https://svn.dealii.org/trunk@6416
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:43:41 +0000 (22:43 +0000)]
Remove
//TODO:[WB,GK] implement Tvmult
since: why should we do it if nobody asked for it in the last two years?
git-svn-id: https://svn.dealii.org/trunk@6415
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:34:30 +0000 (22:34 +0000)]
Fix one more omission.
git-svn-id: https://svn.dealii.org/trunk@6414
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:32:17 +0000 (22:32 +0000)]
Undo gross blunder.
git-svn-id: https://svn.dealii.org/trunk@6413
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:12:31 +0000 (22:12 +0000)]
Remove
//TODO:[WB] (compiler) replace s.c_str() by s when that is possible
This is not realistic: the standard is at it is. It's not a problem of
the compilers.
git-svn-id: https://svn.dealii.org/trunk@6412
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:11:30 +0000 (22:11 +0000)]
Clean up.
git-svn-id: https://svn.dealii.org/trunk@6411
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:06:53 +0000 (22:06 +0000)]
Fix a TODO by moving the exception class to the place it belongs to.
git-svn-id: https://svn.dealii.org/trunk@6410
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 22:03:57 +0000 (22:03 +0000)]
Fix constraint matrices.
git-svn-id: https://svn.dealii.org/trunk@6409
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 21:55:00 +0000 (21:55 +0000)]
Remove seemingly unnecessary exception.
git-svn-id: https://svn.dealii.org/trunk@6408
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 15:48:35 +0000 (15:48 +0000)]
Micro changes for better readability.
git-svn-id: https://svn.dealii.org/trunk@6407
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 15:39:27 +0000 (15:39 +0000)]
Add forgotten includes.
git-svn-id: https://svn.dealii.org/trunk@6406
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 15:30:38 +0000 (15:30 +0000)]
Add missing header includes.
git-svn-id: https://svn.dealii.org/trunk@6405
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 15:08:23 +0000 (15:08 +0000)]
Add missing header includes.
git-svn-id: https://svn.dealii.org/trunk@6404
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 14:48:49 +0000 (14:48 +0000)]
Add missing header includes.
git-svn-id: https://svn.dealii.org/trunk@6403
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 14:44:51 +0000 (14:44 +0000)]
Add missing header includes.
git-svn-id: https://svn.dealii.org/trunk@6402
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Thu, 12 Sep 2002 08:19:47 +0000 (08:19 +0000)]
read past end error removed
git-svn-id: https://svn.dealii.org/trunk@6401
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 00:40:54 +0000 (00:40 +0000)]
Add missing header includes.
git-svn-id: https://svn.dealii.org/trunk@6400
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 00:14:06 +0000 (00:14 +0000)]
Add missing header includes.
git-svn-id: https://svn.dealii.org/trunk@6399
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 00:04:42 +0000 (00:04 +0000)]
Add missing header.
git-svn-id: https://svn.dealii.org/trunk@6398
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 00:03:58 +0000 (00:03 +0000)]
Add missing forward declarations.
git-svn-id: https://svn.dealii.org/trunk@6397
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 00:03:47 +0000 (00:03 +0000)]
Add the all-headers test.
git-svn-id: https://svn.dealii.org/trunk@6396
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 12 Sep 2002 00:01:11 +0000 (00:01 +0000)]
Add missing forward declaration.
git-svn-id: https://svn.dealii.org/trunk@6395
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Wed, 11 Sep 2002 23:58:51 +0000 (23:58 +0000)]
Add missing header.
git-svn-id: https://svn.dealii.org/trunk@6394
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Wed, 11 Sep 2002 20:03:01 +0000 (20:03 +0000)]
Fix missing include file.
git-svn-id: https://svn.dealii.org/trunk@6393
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Wed, 11 Sep 2002 16:06:51 +0000 (16:06 +0000)]
SparseMatrix::const_iterator
git-svn-id: https://svn.dealii.org/trunk@6392
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 15:57:37 +0000 (15:57 +0000)]
Just some cosmetics.
git-svn-id: https://svn.dealii.org/trunk@6391
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 15:50:43 +0000 (15:50 +0000)]
just reindent
git-svn-id: https://svn.dealii.org/trunk@6390
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 15:45:37 +0000 (15:45 +0000)]
Add missing std:: (how annoying that this should happen to _me_!...)
git-svn-id: https://svn.dealii.org/trunk@6389
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 14:52:50 +0000 (14:52 +0000)]
Replace & by &
git-svn-id: https://svn.dealii.org/trunk@6388
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 14:51:54 +0000 (14:51 +0000)]
More short fixes.
git-svn-id: https://svn.dealii.org/trunk@6387
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 14:47:07 +0000 (14:47 +0000)]
One-char typo fix.
git-svn-id: https://svn.dealii.org/trunk@6386
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 14:31:17 +0000 (14:31 +0000)]
More reindentation and timezones.
git-svn-id: https://svn.dealii.org/trunk@6385
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 10 Sep 2002 14:30:22 +0000 (14:30 +0000)]
Reindent, and play a little with timezones.
git-svn-id: https://svn.dealii.org/trunk@6384
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Tue, 10 Sep 2002 08:38:16 +0000 (08:38 +0000)]
new exception ExcIteratorPastEnd
git-svn-id: https://svn.dealii.org/trunk@6383
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 23:46:26 +0000 (23:46 +0000)]
Export DEAL_II_PATH to programs through the preprocessor.
git-svn-id: https://svn.dealii.org/trunk@6382
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 23:37:54 +0000 (23:37 +0000)]
Remove some messing with version numbers that does not seem to work
any more with newer autoconfs.
git-svn-id: https://svn.dealii.org/trunk@6381
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 23:20:36 +0000 (23:20 +0000)]
Remove tmp file again.
git-svn-id: https://svn.dealii.org/trunk@6380
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 23:18:33 +0000 (23:18 +0000)]
SparseMatrix::block_read/write
git-svn-id: https://svn.dealii.org/trunk@6379
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 22:52:47 +0000 (22:52 +0000)]
Use a fixed scaling of the three tableaux, since on my netscape6 the
middle frame doesn't get any height at all...
git-svn-id: https://svn.dealii.org/trunk@6378
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 22:41:08 +0000 (22:41 +0000)]
SparsityPattern::block_write/block_read
git-svn-id: https://svn.dealii.org/trunk@6377
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 21:47:52 +0000 (21:47 +0000)]
Delete erroneous assertion: after reading everything in block_read,
the input stream may not return true after conversion to bool, if we
have reached the end of the file. The assertion was thus wrong.
git-svn-id: https://svn.dealii.org/trunk@6376
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 21:30:55 +0000 (21:30 +0000)]
Reformat some comments.
git-svn-id: https://svn.dealii.org/trunk@6375
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 15:41:35 +0000 (15:41 +0000)]
Fix two incorrectly checked in things.
git-svn-id: https://svn.dealii.org/trunk@6374
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Mon, 9 Sep 2002 14:38:16 +0000 (14:38 +0000)]
Implement n_existing_threads.
git-svn-id: https://svn.dealii.org/trunk@6373
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Mon, 9 Sep 2002 13:14:19 +0000 (13:14 +0000)]
sparse_matrices should be ok
git-svn-id: https://svn.dealii.org/trunk@6372
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Fri, 6 Sep 2002 22:06:34 +0000 (22:06 +0000)]
Remove forward declaration files.
git-svn-id: https://svn.dealii.org/trunk@6371
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Fri, 6 Sep 2002 13:36:49 +0000 (13:36 +0000)]
SparseMatrixEZ const_iterator
git-svn-id: https://svn.dealii.org/trunk@6370
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Fri, 6 Sep 2002 09:08:30 +0000 (09:08 +0000)]
New exception ExcIteratorPastEnd
git-svn-id: https://svn.dealii.org/trunk@6369
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 5 Sep 2002 20:44:26 +0000 (20:44 +0000)]
Lift restriction on use of Threads::PosixThreadBarrier slightly for the case that the respective POSIX functions are not available.
git-svn-id: https://svn.dealii.org/trunk@6368
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 5 Sep 2002 20:27:59 +0000 (20:27 +0000)]
Fix docs.
git-svn-id: https://svn.dealii.org/trunk@6367
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 5 Sep 2002 19:39:36 +0000 (19:39 +0000)]
Work around a problem in MT mode. The docs state what it is.
git-svn-id: https://svn.dealii.org/trunk@6366
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 5 Sep 2002 18:55:57 +0000 (18:55 +0000)]
Remove a syntax error.
git-svn-id: https://svn.dealii.org/trunk@6365
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Thu, 5 Sep 2002 17:54:29 +0000 (17:54 +0000)]
small update
git-svn-id: https://svn.dealii.org/trunk@6364
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Thu, 5 Sep 2002 17:43:24 +0000 (17:43 +0000)]
preconditioners for SparseMatrixEZ
git-svn-id: https://svn.dealii.org/trunk@6363
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 5 Sep 2002 14:55:50 +0000 (14:55 +0000)]
Remove forward declaration files.
git-svn-id: https://svn.dealii.org/trunk@6362
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 5 Sep 2002 14:43:23 +0000 (14:43 +0000)]
Fix a rather nasty bug.
git-svn-id: https://svn.dealii.org/trunk@6361
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Thu, 5 Sep 2002 14:27:47 +0000 (14:27 +0000)]
Check something more.
git-svn-id: https://svn.dealii.org/trunk@6360
0785d39b -7218-0410-832d-
ea1e28bc413d
guido [Wed, 4 Sep 2002 19:44:40 +0000 (19:44 +0000)]
SparseMatrixEZ reprogrammed
git-svn-id: https://svn.dealii.org/trunk@6359
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 17:41:12 +0000 (17:41 +0000)]
Update.
git-svn-id: https://svn.dealii.org/trunk@6358
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 17:39:25 +0000 (17:39 +0000)]
Remove vector2d, use Table.
git-svn-id: https://svn.dealii.org/trunk@6357
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 17:37:44 +0000 (17:37 +0000)]
Spelling fix.
git-svn-id: https://svn.dealii.org/trunk@6356
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 17:27:23 +0000 (17:27 +0000)]
Slightly change output.
git-svn-id: https://svn.dealii.org/trunk@6355
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 17:26:13 +0000 (17:26 +0000)]
More and more thorough testing.
git-svn-id: https://svn.dealii.org/trunk@6354
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 15:36:26 +0000 (15:36 +0000)]
Table<N,T>
git-svn-id: https://svn.dealii.org/trunk@6353
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 15:35:45 +0000 (15:35 +0000)]
Use Table<2,...> instead of vector2d<...>.
git-svn-id: https://svn.dealii.org/trunk@6352
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 15:31:27 +0000 (15:31 +0000)]
Implement generation of gradients of Nedelec elements.
git-svn-id: https://svn.dealii.org/trunk@6351
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 15:30:45 +0000 (15:30 +0000)]
transform_* for tensors of rank 2.
git-svn-id: https://svn.dealii.org/trunk@6350
0785d39b -7218-0410-832d-
ea1e28bc413d
wolf [Tue, 3 Sep 2002 15:30:31 +0000 (15:30 +0000)]
Add and clarify some comments. Add a second set of transform_{co,contra}variant functions that transform matrices instead of vectors.
git-svn-id: https://svn.dealii.org/trunk@6349
0785d39b -7218-0410-832d-
ea1e28bc413d
In the beginning the Universe was created. This has made a lot of
people very angry and has been widely regarded as a bad move.
Douglas Adams
Typeset in
Trocchi and Trocchi Bold Sans Serif.