[phylocom-user] Randomisation in Comdist and Comdistnt

Di Lei D.Lei at student.reading.ac.uk
Wed Aug 17 08:57:30 PDT 2016


Hello everyone,


With this email I would like to ask a question about Comdist and Comdistnt.


If I understood it right, Comdist allows taking sample from the sample itself to generate the NRIrandom, but not in Comdistnt. Thus in the result matrix, samples should cluster with themselves in Comdist, and should not in Comdistnt.


However, in my result, samples did everything with everything with themselves, they clustered, over dispersed or just not related. For example, here is one of my Comdist result, highlight in green is >1.96, in red is <-1.96. Since I can not explain the relationship with sample themselves, I doubt with the result between them.


.       A       B       C       D       E
A       -0.980543       0.569294        -1.870606       3.009378        -2.851256
B       0.566406        3.287449        -0.873764       4.737475        -1.549745
C       -1.870788       -0.873311       -1.516078       1.46536 -2.688112
D       3.009408        4.737756        1.465491        4.850933        1.34407
E       -2.850549       -1.549568       -2.688442       1.343928        -2.235899


Does anyone have an ideal for it?

With kind regards,
Di Lei

MSc student
University of Reading

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.phylodiversity.net/pipermail/phylocom-user-phylodiversity.net/attachments/20160817/74363c37/attachment.htm>


More information about the phylocom-user mailing list