[Networkit] NetworkIt and Xeon Phi

Christian Staudt christian.staudt at kit.edu
Sat Mar 12 09:00:42 CET 2016


> On 12 Mar 2016, at 00:49, Wang, Zhijie <ZHW47 at pitt.edu> wrote:
> 
> Desired accuracy level is low, so any suggested epsilon level or sample size?


I can’t really give you any “optimal” parameters. You could try
	centrality.ApproxBetweenness(G, epsilon=0.1)
	centrality.ApproxBetweenness2(G, nSamples=42)
Turn epsilon up or nSamples down as needed to reduce running time to what you want to invest. If you set the loglevel to INFO
	setLogLevel(“INFO”)
and call centrality.ApproxBetweenness(G, epsilon).run(), it will tell you how many searches are performed, e.g.:

	[INFO ][“networkit/cpp/centrality/ApproxBetweenness.cpp”, 56: virtual void NetworKit::ApproxBetweenness::run()]: taking 831 path samples

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ira.uni-karlsruhe.de/mailman/private/networkit/attachments/20160312/772a7e1a/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://lists.ira.uni-karlsruhe.de/mailman/private/networkit/attachments/20160312/772a7e1a/attachment.sig>


More information about the NetworKit mailing list