CRABS (version 1.2.0)

sample.congruence.class.posterior: Stochastic exploration of congruent models for all samples in the posterior

Description

This function takes a posterior sample as input: a list of CRABS objects. It will then iterate over the samples, and for each posterior sample it will sample from the posterior class. It will sample using the sample.basic.models function, and all additional parameters are passed to sample.basic.models.

Usage

sample.congruence.class.posterior(
  posterior,
  num.samples,
  rate.type = "extinction",
  mu0.equal = FALSE,
  rate0 = NULL,
  ...
)

Value

A named list with congruent rates.

Arguments

posterior

a list of CRABS model objects

num.samples

The number of samples to be drawn

rate.type

either "extinction", "speciation", "both" or "joint"

mu0.equal

whether to propose alternative mu starting at mu0 equal to the posterior sample. default to FALSE

rate0

rate0 allows the user to fix the extinction rate at the present to a single value. defaults to NULL, for drawing it randomly

...

Arguments passed on to sample.basic.models

times

the time knots

model

"MRF" for pure MRF model, otherwise MRF has a trend of type "exponential","linear", or "episodic<n>"

direction

"increase" or "decrease" (measured in past to present)

noisy

If FALSE, no MRF noise is added to the trajectory

MRF.type

"HSMRF" or "GMRF", type for stochastic noise.

monotonic

Whether the curve should be forced to always move in one direction.

fc.mean

Determines the average amount of change when drawing from the model.

rate0.median

When not specified, rate at present is drawn from a lognormal distribution with this median.

rate0.logsd

When not specified, rate at present is drawn from a lognormal distribution with this sd

mrf.sd.scale

scale the sd of the mrf process up or down. defaults to 1.0

min.rate

The minimum rate (rescaling fone after after drawing rates).

max.rate

The maximum rate (rescaling fone after after drawing rates).

Examples

data(primates_ebd_log)

posterior <- read.RevBayes(primates_ebd_log, max_t = 65, n_samples = 10)

samples <- sample.congruence.class.posterior(posterior, 
                                             num.samples = 5,
                                             rate.type = "extinction",
                                             rate0.median = 0.1,
                                             model = "MRF",
                                             max.rate = 1.0)

print(samples)