ci: Compute Confidence Intervals

Description

Compute and display confidence intervals for model estimates. Methods are provided for the mean of a numeric vector ci.default, the probability of a binomial vector ci.binom, and for lm, and lme objects are provided.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
  ci(x, confidence=0.95, alpha=1 - confidence, ...)
  ## S3 method for class 'numeric'
ci(x, confidence=0.95, alpha=1-confidence, na.rm=FALSE, ...)
  ## S3 method for class 'binom'
ci(x, confidence=0.95, alpha=1-confidence, ...)
  ## S3 method for class 'lm'
ci(x, confidence=0.95, alpha=1-confidence, ...)
  ## S3 method for class 'lme'
ci(x, confidence=0.95, alpha=1-confidence, ...)
  
  ## S3 method for class 'estimable'
ci(x, confidence=0.95, alpha=1-confidence, ...)

Arguments

x

object from which to compute confidence intervals.

confidence

confidence level. Defaults to 0.95.

alpha

type one error rate. Defaults to 1.0-confidence

na.rm

boolean indicating whether missing values should be removed. Defaults to FALSE.

...

Arguments for methods

Details

ci.binom computes binomial confidence intervals using the Clopper-Pearson 'exact' method based on the binomial quantile function. Due to the discrete nature of the binomial distribution, this interval is conservative.

Value

vector or matrix with one row per model parameter and elements/columns Estimate, CI lower, CI upper, Std. Error, DF (for lme objects only), and p-value.

Author(s)

Gregory R. Warnes greg@warnes.net

See Also

confint, lm, summary.lm

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
# mean and confidence interval
ci( rnorm(10) )

# binomial proportion and exact confidence interval
b <- rbinom( prob=0.75, size=1, n=20 )
ci.binom(b) # direct call
class(b) <- 'binom'
ci(b)       # indirect call

# confidence intervals for regression parameteres
data(state)
reg  <-  lm(Area ~ Population, data=as.data.frame(state.x77))
ci(reg)

# mer example
library(nlme)
Orthodont$AgeGroup <- gtools::quantcut(Orthodont$age)
fm2 <- lme(distance ~ Sex + AgeGroup, data = Orthodont,random=~1|Subject)
ci(fm2)

Questions? Problems? Suggestions? or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.