summarize.trends {CRABS}

Summarize trends in the congruence class

Description

Summarize trends in the congruence class

Usage

summarize.trends(model_set, threshold = 0.005, rate_name = "lambda", 
window_size = 1, method = "neighbour", per_time = TRUE, return_data = FALSE,
rm_singleton = FALSE, relative_deltas = FALSE, group_names = NULL)

Arguments

model_set

an object of type "CRABSset"

threshold

a threshold for when \Delta \lambda i should be interpreted as decreasing, flat, or increasing

rate_name

either "lambda" or "mu" or "delta"

window_size

the window size "k" in \Delta\lambda i = \lambda i - \lambda(i-k)

method

default to "neighbour", i.e. to compare rate values at neighbouring time points.

per_time

whether to compute \Delta\lambda i that are in units of per time, i.e. divide by \Delta t

return_data

instead of plots, return the plotting dataframes

rm_singleton

whether or not to remove singletons. Pass starting at present, going towards ancient

relative_deltas

whether to divide \Delta \lambda i by the local lambda value

group_names

a vector of prefixes, if you want to group the models in a facet. For example 'c("reference", "model")'

Value

a patchwork object

Examples

data(primates_ebd)
lambda <- approxfun(primates_ebd$time, primates_ebd$lambda)
mu <- approxfun(primates_ebd$time, primates_ebd$mu)
times <- seq(0, max(primates_ebd$time), length.out = 500)

reference <- create.model(lambda, mu, times = times)

mus <- list(function(t) exp(0.01*t) - 0.01*t - 0.9,
            function(t) exp(-0.02*t) - 0.2,
            function(t) exp(-0.07*t) + 0.02*t - 0.5,
            function(t) 0.2 + 0.01*t,
            function(t) 0.2)


model_set <- congruent.models(reference, mus = mus)

p <- summarize.trends(model_set, 0.02)