Welcome to Taniquetil, Eldaran's home!

Well, we are not so high... yet. So welcome at least to homepage of Robert Cimrman.

I am a PhD student of Applied Mechanics at the University of West Bohemia in Plzeň, Czech Republic. My official field of interest is biomechanics, particularly mathematical modelling of smooth muscle tissues. You can look at my (unfinished, unfinnished too) CV.

Some latest (16.07.2001) outputs of work I participate in can be found in results directory.

There is a special campaign here, but it is only for invited. Invited are all who understand this:

Aby si anglicky mluvící moc nevyskakovali, napsal jsem také něco česky. Ať mají zlost, že nerozumí :-). No, pravý důvod pro užití češtiny je ten, že zmíněná kampaň se týká pouze občanů České republiky. Proto občane, následuj odkaz! (To zní ale blbě. Čí odkaz? Lenina? Asi...)

Pokud vás zajímají příklady, které se zatím dělaly na MVS (matematické výpočtové systémy, aka matlab), následujte link. Budou tam i zadání semestrálek. Oficiální stránky jsou tady.

Local Mind Algorithm Projections

This majestic title means that a list follows of all the nonsenses I have coded so far. In fact, all stuff here is a result of the real-work-aversion, too-much-forced-free-time or never-be-adult-approach. I have created as well many tools (including a FEM package) supporting my work, but they are all strictly confidential (read: not finished, buggy and hardly usable for anybody else).

ma2ze is a 3D maze written in the matlab language. Its page is here.
CAFire has its own page here. If you have always dreamt about a burning mouse pointer in your X, this is something for you.
Elwing is a parallel conjugate gradient solver written from scratch in one week. Current implementation uses MPI, but you can use whatever you like - the only thing to do is to write your own "itc.c". Since I was under lack-of-time pressure (never used MPI before), there are some shortcomings or inefficiencies still lurking here. The biggest of them is a memory wasting due to storing both triangles of a symmetrical matrix - the matrix distribution is rather complex even in this case. However it can be rather easily fixed and works quite well: a pentadiagonal system of 1.000.000 equations was solved on one of our university clusters (6 Linux PC on Myrinet) in about 7 seconds. Yes, the matrix was _really_ well conditioned :-), but 3 iterations in less then 1 second is not bad. You can check some performance statistics, or download it and try it yourself.
INM stays for Iterative Numerics Magic. I created this when assisting in Numerical Methods lessons. It displays Newton fractal set and requires Qt library. I wrote it basically to see what the hell is going on in fractal and convergence stuff, that it produces so nice pictures. You can get it here. It has a rather convenient GUI and is not so bad, but if you want something really _good_ and _fast_, get XaoS. It is the best fractal beast I have ever seen.
More comes later, after I fight laziness.

Site Evolution

16.07.2001 The results link added.
22.03.2000 MVS section updated.
10.03.2000 ma2ze moved to its own page.
08.03.2000 MVS link (resources for matlab courses) and a photo from my identification card of KČ-50 Club added.
07.03.2000 ma2ze added.
09.02.2000 CV link added.
17.01.2000 A campaign page added.
14.01.2000 Elwing, INM and a new title added, this list shifted to the end.
11.01.2000 This page was born in great pain and kilolitres of blood.

E-mail: cimrman3 at students dot zcu dot cz