Using Statistics

From BEAST Software

BEAST Documentation->Tutorials-> Using Statistics

How to use statistics to obtain posterior estimates

Sets of taxa can be defined using the <taxa> element. For example, for the dataset defined in tutorial 1, we may wish to define the taxa in the Pan (chimpanzee) genus:

<taxa id="Pan">
	<taxon idref="chimp"/>
	<taxon idref="bonobo"/>
</taxa>

Note the two taxon elements were defined previously and these ones here are references to them (using the idref attribute). You can define as many <taxa> elements as you require using any combination of the available <taxon> elements.

Time of MRCA Statistic

The <tmrcaStatistic> provides the time of the most recent common ancestor (MRCA) of a group of taxa in a tree. Each time this statistic is logged, it finds the node representing the MRCA in the tree and returns the time of that node. This allows a particular divergence date to be recorded event though the tree is changing. By logging this statistic you can obtain a Bayesian posterior dististribution of the divergence of these taxa.

  <tmrcaStatistic name="time_Pan">
  	<treeModel idref="treeModel1"/>
  	<mrca>
  		<taxa idref="Pan"/>
  	</mrca>
  </tmrcaStatistic>

Monophyly Statistic

The <monophylyStatistic> simply returns a value of 1 if the given set of taxa are monophyletic on the given tree and 0 otherwise. The taxa are monophyletic if there is a node in the tree that has as descendants all the taxa in the set and only those taxa.

  <monophylyStatistic name="foo">
  	<treeModel idref="treeModel1"/>
  	<mrca>
  		<taxa idref="Pan"/>
  	</mrca>
  </monophylyStatistic>