More on Heuristic Searches

Dr. Andrew G. McArthur mcarthur at onyx.si.edu
Mon Apr 13 10:36:58 EST 1998


Elisa,

While I do my random replicates heuristic searches with a limit of 20 trees
per replicate, I do my bootstraps (200 reps) with a limit of 10 trees per
replicate.  This is sure to somewhat underestimate the bootstrap support
for nodes, but helps out on speed.  It would be probably better to use the
same restrictions on heuristic and bootstrap searches.  I'm not sure what
you mean by your last sentance - "how would this influence number of
bootstrap replicates?".  If there is a trade-off between number of
bootstrap replicates and number of trees considered in each replicate, I've
no knowledge of how it works.  Hope this helps.

Andrew McArthur
--------------------------------------------
Dr. Andrew G. McArthur, Guest Researcher
University of Western Ontario, Canada
mcarthur at onyx.si.edu, Fax: (519) 439-5981
http://www.geocities.com/CapeCanaveral/8431/
 
At 09:55 AM 4/11/98 -0600, you wrote:
>Hello All,
>
>Here is a new question spinning off from the earlier inquiry and and its
>reply (See Re:heuristic searches). As suggested in the original, when
>your goal is maintain rigorous testing while conducting heuristic
>searches within "reasonable" time frames...If you were to reduce the
>number of trees saved to something like 20 trees in memory for replicate
>before starting another, how would one approach the bootstrap settings?
>Would you simply use the same settings for that second search? And how
>would this influence number of bootstrap replicates?
>
>Working on borrowed computer time does imply a certain necessity of time
>consciousness.  
>
>Best Regards - e2
>
>Elisa Eshbaugh, Graduate Student
>Montana State University
>Bozeman, MT
>Eshbaugh at avicom.net
>
>
>




More information about the Mol-evol mailing list