extractAIC.regimix

Computes the generalised AIC for a regimix model.

misc

Identifies regions of common (species) profiles (RCPs), possibly when sampling artefacts are present. Within a region the probability of sampling all species remains approximately constant. This is performed using mixtures-of-experts models. The package also contains associated methods, such as diagnostics. Details of the method can be found in Foster et al (2013) <doi:10.1002/env.2245> and Foster et al. (2017) <doi:10.1111/rssc.12211>.

Scott Foster

RCPmod

Regions of Common Profiles Modelling with Mixtures-of-Experts

extractAIC.regimix function

extractAIC( object, scale, k=2, ...)

Method

<dl><dt> object</dt>
<dd>an x obtained from fitting a region of common profile mixture model. Such as that generated from a call to regimix(qv).</dd><dt> scale</dt>
<dd>ignored</dd><dt> k</dt>
<dd>the coefficient for the penalty in the information criterion. k=2 corresponds to Akaikes information criterion, k=log(x$n) corresponds to the Bayesian information criterion. Default is k=2 (AIC).</dd><dt> ...</dt>
<dd>ignored</dd></dl>

Arguments

Author

Extracts the AIC for a regimix model. — extractAIC.regimix

<dl>

<dt> object</dt>
<dd>an x obtained from fitting a region of common profile mixture model. Such as that generated from a call to regimix(qv).</dd>

<dt> scale</dt>
<dd>ignored</dd>

<dt> k</dt>
<dd>the coefficient for the penalty in the information criterion. k=2 corresponds to Akaikes information criterion, k=log(x$n) corresponds to the Bayesian information criterion. Default is k=2 (AIC).</dd>

<dt> ...</dt>
<dd>ignored</dd>

</dl>

extractAIC.regimix: Extracts the AIC for a regimix model.

Description

Value

Method

Arguments

Author