<html>
<head>
<style>
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Verdana
}
</style>
</head>
<body class='hmmessage'>Yes it helps since I understood there is no bug here, although I still think it would be more useful for users to be able to detect when the integration has gone out of the domain, so as to stop integration one step earlier.<br>Best regards<br><br><hr>Date: Tue, 18 Aug 2009 15:27:15 +0200<br>From: jbiddiscombe@skippingmouse.co.uk<br>To: stan1313@hotmail.fr<br>CC: zhanping.liu@kitware.com; biddisco@cscs.ch; paraview@paraview.org<br>Subject: Re: [Paraview] RungeKutta4 implementation issue<br><br>
<br>
<blockquote cite="mid:BAY109-W3569904C1BF8F015FC4BAFB2FF0@phx.gbl">My main concern is why does the integrator return 0
instead of OUT_OF_DOMAIN?<br>
</blockquote>
<br>
Because the (current position inside domain + computed velocity * time)
is returned - this final result is out of domain as you noticed, but
until you pass that new position value back in and ask for the next
point, it doesn't check that the value is in/out.<br>
<br>
It checks that the start point, and all intermediate computed points
are inside the domain, but if the final result pops out - then it means
that the final step was too big (or the data was a bit rubbish -
typically when every Nth time step is saved - the courant condition is
no longer satisfied etc etc).<br>
<br>
does that help?<br>
<br>
JB<br>
<br>
<br>
<blockquote cite="mid:BAY109-W3569904C1BF8F015FC4BAFB2FF0@phx.gbl">Regards<br>
<br>
<br>
<hr>Date: Mon, 17 Aug 2009 10:45:31 -0400<br>
Subject: Re: [Paraview] RungeKutta4 implementation issue<br>
From: <a class=EC_moz-txt-link-abbreviated href="mailto:zhanping.liu@kitware.com">zhanping.liu@kitware.com</a><br>
To: <a class=EC_moz-txt-link-abbreviated href="mailto:biddisco@cscs.ch">biddisco@cscs.ch</a><br>
CC: <a class=EC_moz-txt-link-abbreviated href="mailto:stan1313@hotmail.fr">stan1313@hotmail.fr</a>; <a class=EC_moz-txt-link-abbreviated href="mailto:paraview@paraview.org">paraview@paraview.org</a><br>
<br>
<div>Hi Fred:</div>
<div> </div>
<div> As JB suggested, please try using RK45, which employs an
adaptive step size that is insensitive to the velocity magnitude and
hence is capable of capturing accurate shape of the curve (flow line).
Using RK4 (still adopting a fixed step size) does not guarantee a
sufficient / desired acurracy of numerical integration (though
certainly it is better than RK2) since it is stil sensitive to the
velocity magnitude.</div>
<div> </div>
<div> Recently we improved the flow line integration accuracy
issue. For any problems, please feel free to let us know.</div>
<div> </div>
<div> Thanks.</div>
<div> </div>
<div> -Zhanping<br>
</div>
<div> -- <br>
Zhanping Liu, PhD<br>
Kitware, Inc.<br>
28 Corporate Drive<br>
Clifton Park, NY 12065-8662<br>
Phone: 518-371-3971 x 138<br>
<a href="http://www.zhanpingliu.org" target="_blank">http://www.zhanpingliu.org</a><br>
<br>
</div>
<div class="EC_EC_gmail_quote">On Mon, Aug 17, 2009 at 5:06 AM, John
Biddiscombe <span dir=ltr><<a href="mailto:biddisco@cscs.ch">biddisco@cscs.ch</a>></span> wrote:<br>
<blockquote class="EC_EC_gmail_quote" style="padding-left:1ex">
<div>I suggest you attach a debugger to the code and step through.
Then you'll be able to answer your question. <br>
<br>
Since the velocity field of your data may be high, it is entirely
possible that the computed next position is outside of the bounds.
Perhaps reducing the step size will help. Try even Rk4.5 with adaptive
steps, this may help if the velocity gradient at the edge of the domain
is very large..<br>
<br>
JB<br>
<br>
<blockquote>
<div>
<div class=h5>How is it possible???<br>
<br>
double *bds = reader->GetOutput()->GetBounds();<br>
printf("%lf %lf %lf %lf\n", bds[0], bds[1], bds[2], bds[3]);<br>
<br>
-> 0.000000 69.000000 0.000000 69.000000<br>
<br>
printf("point1[0]=%f point1[1]=%f point1[2]=%f\n", point1[0],
point1[1], point1[2]);<br>
<br>
-> point1[0]=58.196274 point1[1]=68.862898 point1[2]=0.000000<br>
<br>
if (integrator->ComputeNextStep(point1, point2, 0, delT.Interval,
stepTaken, minStep, maxStep, this->MaximumError, error) != 0)<br>
break;<br>
printf("point2[0]=%f point2[1]=%f point2[2]=%f\n", point2[0],
point2[1], point2[2]);<br>
<br>
-> point2[0]=59.002164 point2[1]=69.454414 point2[2]=0.000000<br>
<br>
PS: This message should probably be sent to the VTK mailing list but it
seems that it is no longer so active...<br>
<br>
<br>
<hr>
Découvrez toutes les possibilités de communication <a href="http://www.microsoft.com/windows/windowslive/default.aspx" target="_blank">avec vos proches</a> </div>
</div>
<pre><hr size=4 width="90%">
_______________________________________________
Powered by <a href="http://www.kitware.com/" target="_blank">www.kitware.com</a>
Visit other Kitware open-source projects at <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a>
Please keep messages on-topic and check the ParaView Wiki at: <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a>
Follow this link to subscribe/unsubscribe:
<a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a>
</pre>
</blockquote>
<br>
<font color="#888888"><br>
<pre>--
John Biddiscombe, email:biddisco @ <a href="http://cscs.ch/" target="_blank">cscs.ch</a>
<a href="http://www.cscs.ch/" target="_blank">http://www.cscs.ch/</a>
CSCS, Swiss National Supercomputing Centre | Tel: +41 (91) 610.82.07
Via Cantonale, 6928 Manno, Switzerland | Fax: +41 (91) 610.82.82</pre>
</font></div>
<br>
_______________________________________________<br>
Powered by <a href="http://www.kitware.com/" target="_blank">www.kitware.com</a><br>
<br>
Visit other Kitware open-source projects at <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
<br>
Please keep messages on-topic and check the ParaView Wiki at: <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a><br>
<br>
Follow this link to subscribe/unsubscribe:<br>
<a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a><br>
<br>
</blockquote>
</div>
<br>
<br clear=all>
<br>
<br>
<hr>Vous voulez savoir ce que vous pouvez faire avec le nouveau
Windows Live ? <a href="http://www.microsoft.com/windows/windowslive/default.aspx" target="_blank">Lancez-vous !</a>
<pre>
<hr size=4 width="90%">
_______________________________________________
Powered by <a class=EC_moz-txt-link-abbreviated href="http://www.kitware.com" target="_blank">www.kitware.com</a>
Visit other Kitware open-source projects at <a class=EC_moz-txt-link-freetext href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a>
Please keep messages on-topic and check the ParaView Wiki at: <a class=EC_moz-txt-link-freetext href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a>
Follow this link to subscribe/unsubscribe:
<a class=EC_moz-txt-link-freetext href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a>
</pre>
</blockquote>
<br /><hr />Discutez sur Messenger où que vous soyez ! <a href='http://www.messengersurvotremobile.com/' target='_new'>Mettez Messenger sur votre mobile !</a></body>
</html>