<html dir="ltr">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<style>
<!--
.hmmessage p
{margin:0px;
padding:0px}
body.hmmessage
{font-size:10pt;
font-family:Tahoma}
-->
</style><style id="owaParaStyle" type="text/css"></style>
</head>
<body ocsi="0" fpstyle="1" class="hmmessage">
<div style="direction: ltr;font-family: Tahoma;color: #000000;font-size: 10pt;">Hi there,
<br>
<br>
I would not call Evanno et al's approach 'formal', but anyway are approach with find.cluster is fairly ad hoc as well. The idea, somewhat arbitrary, is to look for an elbow in the curve of BIC values. Please have a look at the DAPC paper for examples.
<br>
<br>
I developed some 'objective' criteria for automated selection of K, but none if these options is 'better' than visual inspection of the data (see ?find.clusters from the devel version).<br>
<br>
Cheers<br>
<br>
Thibaut.<br>
<div style="font-family: Times New Roman; color: rgb(0, 0, 0); font-size: 16px;">
<hr tabindex="-1">
<div style="direction: ltr;" id="divRpF373203"><font size="2" color="#000000" face="Tahoma"><b>From:</b> Martin Llewellyn [llewellynmartin@hotmail.com]<br>
<b>Sent:</b> 23 May 2011 12:02<br>
<b>To:</b> Jombart, Thibaut; adegenet-forum@r-forge.wu-wien.ac.at<br>
<b>Subject:</b> RE: [adegenet-forum] question find.clusters and BIC values<br>
</font><br>
</div>
<div></div>
<div>Dear Thimbaut and adegenet forum,<br>
<br>
Is there a formal means of identifying the appropriate number of clusters from the BIC value curve if no asymptope is reached? Structure users generally use delta-K (Evanno et al 2005) - would an exquivalent statistic be approrpiate ?<br>
<br>
Thanks<br>
<br>
M<br>
<br>
<hr id="stopSpelling">
From: t.jombart@imperial.ac.uk<br>
To: mrjonker@gmail.com; adegenet-forum@r-forge.wu-wien.ac.at<br>
Date: Thu, 19 May 2011 08:40:45 +0000<br>
Subject: Re: [adegenet-forum] question scatter.dapc<br>
<br>
<style id="ecxowaParaStyle">
<!--
.ExternalClass p
{margin-bottom:0}
-->
BODY {direction: ltr;font-family: Tahoma;color: #000000;font-size: 10pt;}P {margin-top:0;margin-bottom:0;}</style>
<div style="direction: ltr; font-family: Tahoma; color: rgb(0, 0, 0); font-size: 10pt;">
<br>
Dear Rudy, <br>
<br>
the "d=..." simply indicates the mesh of the grid, which is not necessarily useful for a single graph, but may be used to compare different graphs (e.g. plane 1-2 and 3-4, to see how the distances in one plot relate to the other one). You can remove this indication
using "cgrid=0" or remove the grid using "grid=FALSE". <br>
<br>
About the inset, this is generated by the function add.scatter.eig. It scales so that the largest value uses most of the inset window. For instance:<br>
plot(1:10)<br>
add.scatter.eig(10:1,2,1,2)<br>
add.scatter.eig(1:10,0,9,10,posi="bottomright")<br>
<br>
About the inertia ellipses, they are graphical summaries of a cloud of points. This topic is actually being discussed in the adelist forum (forum for ade4):<br>
<a href="http://listes.univ-lyon1.fr/wws/arc/adelist/2011-05/msg00003.html" target="_blank">http://listes.univ-lyon1.fr/wws/arc/adelist/2011-05/msg00003.html</a><br>
<br>
They're not generally confidence ellipses. They are whenever the cloud of point is a sample from a bivariate normal distribution. In this case, "cellipse" (determines the size of the ellipse) indirectly controls the alpha threshold. 1.5 (default) corresponds
to 67%; 2.5 corresponds to the magical 95%.<br>
The exact formula is:<br>
p = 1 − exp( − cellipse^2 /2 )<br>
<br>
<br>
All the best<br>
<br>
Thibaut<br>
<br>
<font size="2"></font>
<div>
<div class="ecxBodyFragment"></div>
</div>
<div style="font-family: Times New Roman; color: rgb(0, 0, 0); font-size: 16px;">
<hr tabindex="-1">
<div id="ecxdivRpF75495" style="direction: ltr;"><font size="2" color="#000000" face="Tahoma"><b>From:</b> adegenet-forum-bounces@r-forge.wu-wien.ac.at [adegenet-forum-bounces@r-forge.wu-wien.ac.at] on behalf of Rudy Jonker [mrjonker@gmail.com]<br>
<b>Sent:</b> 18 May 2011 09:39<br>
<b>To:</b> adegenet-forum@r-forge.wu-wien.ac.at<br>
<b>Subject:</b> [adegenet-forum] question scatter.dapc<br>
</font><br>
</div>
<div></div>
<div>
<div>Dear Thibaut and group,</div>
<div> </div>
<div>I have a question with respect to the graphs made with <em>scatter.dapc</em>. I noticed a d=<em>n</em> in the top right corner of the graph and was wondering what it stands for. I could not find it in the manual or in the article. I also have a question
with respect to the eigenvalue inset in the graph. Are the bars scaled such that the first eigenvalue is always the full length of the graph and that the rest is then plotted relative to the first?
</div>
<div> </div>
<div>And then the inertia ellipse: is this a sort of 95% confidence interval on the position of the centroid of a cluster?</div>
<div> </div>
<div>Thanks in advance,</div>
<div>Rudy</div>
</div>
</div>
</div>
<br>
_______________________________________________ adegenet-forum mailing list adegenet-forum@lists.r-forge.r-project.org https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/adegenet-forum
</div>
</div>
</div>
</body>
</html>