diversitree (version 0.9-3)

time: Support Functions for Time-varying Models

Description

Utility functions for working with time-dependent models.

Usage

constant.t(t, c)
linear.t(t, c, m)
exponential.t(t, y0, y1, r)
sigmoid.t(t, y0, y1, tmid, r)
stepf.t(t, y0, y1, tc)
make.pars.t(functions, cache=NULL, check.negative.const=TRUE,
            check.negative.var=TRUE)

Arguments

t

Time

c

Constant. For linear.t this is the value at t=0.

m

Slope

y0

y value at time 0 (exp.t) or very early times (limit as t tends to negative infinity in sigmoid.t)

y1

y value approached as t tends to negative infinity

r

Rate at which exponential decay occurs or sigmoid changes - higher values are steeper

tmid

Midpoint (inflection point) of sigmoid

tc

"Critical time" at which y changes from y0 to {y1}.

functions

A list of functions with t as their first argument.

cache

Internal use -- please leave NULL.

check.negative.const, check.negative.var

Logicals, indicating if negative parameters should be checked for, for time-constant and time-varying parameters respectively.

Details

The linear function (linear.t) will go to zero wherever negative. This may not always be desired, but is often useful for models of rates.