Talk:Computer simulation/Archive 1

Half the story
It is said that computer simulations build on, and are a replacement for mathematical models. In my opininion, this only tells half of the story. Computer simulations also build on and are replacements for experiments. E.g., experiements in silico are used a lot in biology, and biology is not very much mathematized. In this case, the experiment is more a replacement of traditional experimentation than a replacement of mathematical modelizing. However, I agree it needs a modelization phase. Well, to summarize, IMHO the current article reflects only one side of the story: the modeling apsect (similar to maths) and totally avoids the experimentation aspect of computer simulations. Should be completed to aknowledge both aspects... See also: experiements in silico--Powo 14:24, 1 December 2005 (UTC)

Correction
It is not true that Richard Feynman considered computer models to be a "disease" ... this is a fallacy that was written by Michael Crichton and is oft-repeated by global warming deniers

What Feynman said was: "There is a computer disease that anybody who works with computers knows about. It's a very serious disease and it interferes completely with the work. The trouble with computers is that you 'play' with them!"

Feynman was a pioneer of computer modeling

Java-man 06:59, 31 July 2006 (UTC)

proposal
There could be a "simulation software comparison list" for "commercial product" and "open source".

Recently I did some study in simulation software(mostly commercial product). It is quite few comparison for different simulation softwares. An example is here. But there are still many software not on the list.

And I think this is good evaluation criteria.

is this proposal practicable? There so many major simulation software, and many information can not be easily found on internet.

84.162.77.95 17:16, 19 December 2006 (UTC)


 * This proposal seems somewhat unwieldy. It seems as though the focus of this page should be on the theory and history of computer simulation.  The list of viable software simulations could be endless and is certainly an invitation for commercial exploitation of this page.  Regards. Anlace 19:50, 19 December 2006 (UTC)

I agree that the above proposal, while meritorious in principle, is unwieldy in practice. Sadly, I am the culprit who introduced some of the External links, including the Tool links for what I thought were significant and widely used commercial tools. Since that time a few, let's say, "less-significant" tools have crept in. We are therefore potentially embarked on the slippery slope of commercial exploitation. Since there are so many tools on the market for a wide variety of needs, the solution may be to only include external links to "good" third party tool survey pages. This would provide some modicum of utility to the prospective reader, while avoiding direct commercialization. Alternatively, we may wish to consider eliminating the Tools portion of the external links altogether. I'm not sure which is the best way to go. Any thoughts? ThreePD 00:38, 21 December 2006 (UTC)

Computer?
Not sure what the computer does in the name of this lemma. It is just a tool that makes simulation easier to perform, and nothing more. The lemma should simply be Simulation, imho.--Guido den Broeder 14:56, 27 March 2007 (UTC)

I agree with you. I think the more appropriate term these days is simulation software.

Carl142 23:40, 1 May 2007 (UTC)
 * 22-Oct-2007: Formerly, computers mostly ran the simple math models, but for over 35 years, many, many computer simulations have grown to become massive simulations, where the "tool" has come to life and evolved into generations of living toolisms. These days, saying a "computer is just a tool" (for simulation) is like saying a spaceship is just a tool for travelling, like a "very-fast horse with a heatshield"; i.e, a gross oversimplification.  Ergo, this "lemma" has devolved into a myriad of dilemmas: how to describe the quickly exploding worldwide diversity and numerous current applications of gargantuan computer simulations, within the scope of just a few dozen articles.  Modeling with advanced mathematics was exceeded decades ago, and computer simulation has grown way beyond mathematics. -Wikid77 13:30, 22 October 2007 (UTC)

Rename this article?
I think that the term computer modeling would be a more appropriate term for this article, and the requested article to be merged should be left as a separate article called simulation software. The renamed computer modeling article addresses some very important issues, and the new simulation software article brings additional insight to the subject area. These terms mean different things to different people and merging them may create some confusion or misperception. Carl142 23:52, 1 May 2007 (UTC)
 * I agree. Thegreatdr 17:16, 10 May 2007 (UTC)
 * I agree that rename to Computer modeling is appropriate. Not in favour of merger with Simulation software at this time. Anlace 17:20, 10 May 2007 (UTC)
 * Hmm, no, this would create a similar confusion. In modeling, too, a computer is just a tool. Guido den Broeder 19:23, 2 July 2007 (UTC)

Pitfalls in computer simulation section is factually incorrect
The "pitfalls in computer simulation" section is incorrect; if I have the height of 1,000,000 people, correct to 2 s.f, I can produce a confidence interval for the mean height of the people, accurate to more than 2 s.f, making only the assumption that the people's heights are randomly distributed with a continuous distribution.

Equally, a number might be used in a calculation, known to one significant figure; but if that number is scaled to have only a small effect on the final result, then the final result might well be much more accurate than 1 s.f. - for example, if I know that a person's height it 2m, to 1 s.f, and I know that they are standing on a building that is exactly 100m heigh, then I know the height to the top of their head is 102m, to 3 sf.

The "rules of significance arithmetic" are rules of thumb; this is why sensitivity analysis is used in modelling! - Pog 13:30, 3 August 2007 (UTC)

Have modified the article to reflect the above - Pog 11:12, 6 August 2007 (UTC)

Leader
The opening section of this article can be strengthened, including the use of concepts such as symbolic processing. Will get back to it later, if nobody else beats me to it. So much to do, so little time. -- Iterator12n   Talk  03:30, 5 December 2007 (UTC)

Merger proposal 2
I propose that Computer simulator be merged into this page. The two pages have precisely the same scope; this page is far more complete. I invite someone familiar with simulation to perform the merge. Thanks, WalterGR (talk) 09:48, 3 January 2008 (UTC)


 * It seems to me that Computer simulator is really a very different topic. If anything, the few hints in this page (Computer simulation) that seem to relate to simulators should be moved to the other page.  Simulation as (mostly) discussed here is a fundamentally different topic.


 * Perhaps that would be a better development as scientific simulation (the focus of this page) is a large topic in its own right. Simulators may be a big topic, too, but I can't speak to that.  What I can confirm is that with a shelf full of reliable sources sitting in my office, none of these books mention simulators at all.  So they seem different enough to warrant two different articles to me.  If there is general disagreement about that, then the separation of the topics should me made clear by a redesign of this page. 128.195.89.186 (talk) 21:34, 19 February 2008 (UTC)

Adding a Link into Examples section
After a discussion here - I would like to suggest following link for inclusion into the Examples section: CryoVision - 3D Cryoablation Simulation On my opinion this example (on the contrary to many current examples) provides a short but full description of a case of modern application of computer simulation. The text is understandable and, I hope, interesting for general public. Furthermore nice 3D images are also included. If somebody objects to the inclusion, please let me know. —Preceding unsigned comment added by 79.178.2.247 (talk) 11:52, 15 February 2011 (UTC)
 * I've already noted my view in the edit history: "promotional link to ... software company". This is a spam link, but links normally to be avoided, particularly counts #1, #4 and #5, also apply.


 * The Wikipedia version of "two wrongs don't make a right" is here. A more brutal version of the policy is "don't add sewage to the already polluted pond", here.
 * --Old Moonraker (talk) 12:26, 15 February 2011 (UTC)


 * " but links normally to be avoided, particularly counts #1, #4 and #5, also apply". re #1 - I mentioned already what is unique about our site, this should also cover these remarks: "The Wikipedia version of "two wrongs don't make a right" is here. A more brutal version of the policy is "don't add sewage to the already polluted pond", here". re #4 and #5 - because of the nature of our services we do not expect our clients to come to us through Wikipedia, at least not directly. People who do not know what a computer simulation is - probably would not hire us, I hope its clear. Furthermore, the particular page in question, contains no advertisement (one has to move to home page or to the services page). —Preceding unsigned comment added by 79.178.2.247 (talk) 13:00, 15 February 2011 (UTC)
 * ps: one can publish only the above mentioned link, without link on the company's home page. —Preceding unsigned comment added by 79.178.2.247 (talk) 13:13, 15 February 2011 (UTC)


 * After several days of discussions on different places, it turns out that Old Moonraker, who has initially deleted the link without actually seeing the web site, is the only person who objects to the link inclusion. After having responded to all his comments, with your permission, I restore the deleted link removing the link on our home page. —Preceding unsigned comment added by 79.178.2.247 (talk) 09:55, 16 February 2011 (UTC)
 * Just because I'm the only other editor bothering to comment doesn't make a case for its not being WP:SPAM and WP:ELNO. It is. Deleted. --Old Moonraker (talk) 08:35, 17 February 2011 (UTC)


 * It's a kind of a non democratic behavior. You are not the only one "who cares" here in Wikipedia. If others do not comment it means they do not think the case is WP:SPAM and WP:ELNO. In your discussion you provide no arguments. Let me explain you once again: the link is not WP:SPAM because R&D directors (who are our potential clients) usually know what a computer simulation is, those who read the article can't be our clients. I understand that you used to fight spam here, but you should be a bit more open minded and see that sometimes web site is not spam even though it has commercial background. Furthermore this link brings value to the article(already mentioned several times - see above) that's why its not just another example, its better than the existing ones, and not WP:ELNO. I started this discussion for some reason, so, please Old Moonraker, let others speak, or at least provide some reasons for you deeds.
 * Hi 79.178.2.247. Interest in this discussion seems low indeed, which does not automatically mean people approve. I don't see any added value behind this external link and the commercial factor is certainly an important argument not to include the link. I should add non of the external links in article meet our criteria so I have removed the lot of them. Yoenit (talk) 10:17, 17 February 2011 (UTC)
 * The discussion over at WikiProject Spam came to a similar conclusion. --Hu12 (talk) 19:05, 18 February 2011 (UTC)

Re Grey box completion and validation
“See also“ “Grey box completion and validation“ has been removed from this and several other topics. Following advice from Wikipedia if there are no objections (please provide your name and reasons), I plan to reinstate the reference in a weeks time.

The removed reference provides techniques for computer model development as is needed for many computer simulations. In particular most models are incomplete (i.e. a grey box) and thus need completion and validation. This reference seems to be within the appropriate content of the “See also” section see Wikipedia:Manual_of_Style/Layout#See_also_section. BillWhiten (talk) 09:31, 4 April 2015 (UTC)

Reinstated SolidPhase 5 April 2015 thanks BillWhiten (talk) 05:58, 11 April 2015 (UTC)

according to the book, you can never know the degree of elasticity
Some excerpts :

struct { int elec_or_prot; float x,y,z; /* coordinates */ float xv,yv,zv; /* vector components */ } atom[num_of_atoms];

while (1 ) { /* receive the forces affecting each atom */

n=0; while ( n < num_of_atoms ) {  x_buf2=0, y_buf2=0, z_buf2=0; x_add=0; y_add=0; z_add=0; n2=0; while ( n2 < num_of_atoms ) {   if ( n2==n ) n2++; if ( n2 >= num_of_atoms ) break; power= sqrt (    ( atom[n].x - atom[n2].x)  *    ( atom[n].x - atom[n2].x )+       ( atom[n].y - atom[n2].y) *    ( atom[n].y - atom[n2].y)+     ( atom[n].z - atom[n2].z) *    ( atom[n].z - atom[n2].z) ); old_power=power; /* the distance of the current atom from the foreign         atom (3D-Pythagoras)*/

x_buf=(atom[n].x - atom[n2].x); /* distance components */ y_buf=(atom[n].y - atom[n2].y); z_buf=(atom[n].z - atom[n2].z); if ( power-SCHEITEL < 0.0 ) sign= -1.0; else sign= 1.0; power=power-SCHEITEL; /* there's a ball around each atom and on it's edge the direction of the force is inverted */

if ( fabs(power) < 0.01 ) power=0.01*sign;

power= ( FORCE / ((power*power)+MAXPOWER) *sign ) ;

if ( fabs ( power) >= FORCENULL ) power=0; /* calculate the force that results from the distance */

if ( old_power != 0 ) /* part the force on x, y, and z components */ {         y_buf*=power/old_power; x_buf*=power/old_power; z_buf*=power/old_power;

} else x_buf=0, y_buf=0, z_buf=0;

if ( ( atom[n].elec_or_prot==PROT && atom[n2].elec_or_prot==ELEK ) ||       ( atom[n].elec_or_prot==ELEK && atom[n2].elec_or_prot==PROT )       ) {  y_buf*=-1.0; x_buf*=-1.0; z_buf*=-1.0; } /* opposed forces pull together, same forces push away */

if ( (old_power=sqrt ( x_buf*x_buf+y_buf*y_buf +z_buf*z_buf)) > TRAEGSCHWELLE ) {     power= (old_power-TRAEGSCHWELLE)/ old_power; x_buf*=power; y_buf*=power; z_buf*=power; } else x_buf=0, y_buf=0, z_buf=0;  /* the vector is only changed, when the force has a certain minimum of strength */

if ( old_power > 0 && old_power > TRAEGSCHWELLE) {     x_buf2+=(x_buf*ELANLOSS/TIMESLICE); y_buf2+=(y_buf*ELANLOSS/TIMESLICE); z_buf2+=(z_buf*ELANLOSS/TIMESLICE); } /* change coordinates in force direction so the force isn't added to         the vector. This is tha actual problem. */

x_add+=x_buf; y_add+=y_buf; z_add+=z_buf;

n2++; }

atom[n].xv+=x_add; /* now change the vector components */ atom[n].yv+=y_add; atom[n].zv+=z_add;

atom[n].x+=x_buf2, atom[n].y+=y_buf2, atom[n].z+=z_buf2;

n++; }

n=0;

while ( n < num_of_atoms ) {  atom[n].x+=atom[n].xv/TIMESLICE; atom[n].y+=atom[n].yv/TIMESLICE ; atom[n].z+=atom[n].zv/TIMESLICE;

n++; /* move the atoms according to their vectors */ }

}

That's what i made up. It's still not at an usable stage, it only simulates some collision conditions, but google for "QT Elastic Nodes Example" and you will find that the professional, working solution isn't much different from that. If you would calculate the rotation and position of a graphical depiction of a racing car from such an atom grid, for example, it would be even more impressive.

But the real problem is elasticity. In an unelastic collision, a part of the energy gets lost and is transformed into light (which then causes heat). And you can perhaps never know the amount of energy that gets lost because this also depends on the collision speed or depth. Think about it. If this does not work correctly with an algorithm, it is not possible to prove anything at all. Then not only a computer simulation would not work. — Preceding unsigned comment added by 87.143.84.134 (talk) 21:03, 9 May 2015 (UTC)

Distracting animating image
the animating image is very distracting while trying to read the text of the article, please remove the animation. Sofia Koutsouveli (talk) 13:14, 23 November 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Computer simulation. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20110721023918/http://pem.utbm.fr/materiaux_2002/file/pdf/AF01078.PDF to http://pem.utbm.fr/materiaux_2002/file/pdf/AF01078.PDF

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 21:18, 11 August 2017 (UTC)