RatePhylo.allCI: Confidence intervals for rate parameters

Description Usage Arguments Value Author(s) References Examples

View source: R/RatePhylo.allCI.R

Description

Calculates approximate confidence intervals for all rate parameters. CIs are esimated for one rate parameters while fixing others at a given value (usually the maximum likelihood estimate).

These are reliable (given the asympotic assumptions of the chi-square distribution) if only two groups are being compared but should be regarded only as a rough approximation for =>3 different rate categories. If the rates are correlated the CIs may be underestimated.

Usage

1
RatePhylo.allCI(rateData, MLrate = NULL, fixed = NULL, rateMIN = 0.001, rateMAX = 50, common.mean = FALSE, lambda.est=TRUE)

Arguments

rateData

an object of class "rateData"

MLrate

a vector of relative rate parameters. The length of the vector is equal to the number of rates being estimated. If rate=NULL then rates are equal. Normally these will be the maximum likelihood rate estimates.

fixed

A vector stating whether each parameter should be allowed to vary (either FALSE which results in a start value of 1, or a numeric start value) or should be fixed (TRUE).

rateMIN

Minimum value for the rate parameters

rateMAX

Maximum value for the rate parameters

common.mean

a logical specififying whether each rate category should have its own mean (common.mean=FALSE) or all categories should have the same mean (common.mean=FALSE). See Thomas et al. (2009) for a discussion on the impact of assumptions about mean on rate estimates.

lambda.est

Logical. Estimate Pagel's lambda.

Value

rateLci

Lower confidence interval for rate estimate

rateUci

Upper confidence interval for rate estimate

Author(s)

Gavin Thomas, Rob Freckleton

References

Thomas GH, Freckleton RP, & Szekely T. 2006. Comparative analyses of the influence of developmental mode on phenotypic diversification rates in shorebirds. Proceedings of the Royal Society B 273, 1619-1624.

Thomas GH, Meiri S, & Phillimore AB. 2009. Body size diversification in Anolis: novel environments and island effects. Evolution 63, 2017-2030.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
data(anolis.tree)
data(anolis.data)

## Convert data to class rateData with a rateMatrix object as input
anolis.rateMatrix <- as.rateMatrix(phy=anolis.tree, x="geo_ecomorph", data=anolis.data)

anolis.rateData <- as.rateData(y="Female_SVL", x="geo_ecomorph", 
rateMatrix = anolis.rateMatrix, phy=NULL, data=anolis.data, log.y=TRUE)

# A model with a different rate in each of the four groups. The 'fixed' command is used to determine
# whether a particular rate is to be constrained or not. Use '1' to fix a group and 'FALSE' to show 
# that the parameter is not fixed and should be estimated. The values should be entered in the same
# order as the ranking of the groups. That is, group 0 (small islands) takes position one in the 
# fixed vector, group 1 (large island trunk crown and trunk ground) takes position 2 and so on. The 
# default is to allow each group to take a different mean. 

# Not run
#anole.ML <- optim.likRatePhylo(rateData=anolis.rateData, rate=NULL, #fixed=c(FALSE,FALSE,FALSE, FALSE),
#common.mean=FALSE) # ML estimates

# Confidence intervals for each parameter in turn
# Not run
#RatePhylo.CI(rateData=anolis.rateData, MLrate = anole.ML$MLRate, fixed=c(FALSE, #TRUE, TRUE, TRUE), 
#rateMIN = 0.001, rateMAX = 50, common.mean = FALSE)

#RatePhylo.CI(rateData=anolis.rateData, MLrate = anole.ML$MLRate, fixed=c(TRUE, #FALSE, TRUE, TRUE), 
#rateMIN = 0.001, rateMAX = 50, common.mean = FALSE)

#RatePhylo.CI(rateData=anolis.rateData, MLrate = anole.ML$MLRate, fixed=c(TRUE, #TRUE, FALSE, TRUE), 
#rateMIN = 0.001, rateMAX = 50, common.mean = FALSE)

#RatePhylo.CI(rateData=anolis.rateData, MLrate = anole.ML$MLRate, fixed=c(TRUE, #TRUE, FALSE, TRUE), 
#rateMIN = 0.001, rateMAX = 50, common.mean = FALSE)

# All confidence intervals together
#RatePhylo.allCI(rateData=anolis.rateData, MLrate = anole.ML$MLRate, fixed = #NULL, rateMIN = 0.001, 
#rateMAX = 50, common.mean = FALSE) 

ghthomas/motmot documentation built on July 31, 2018, 12:11 a.m.