ci_int: Construct confidence intervals in the presence of...

In pbreheny/breheny: Miscellaneous Functions

Construct confidence intervals in the presence of interactions

Description

Construct confidence intervals in the presence of interactions

Usage

ci_int(fit, term, ...)

Arguments

fit	A model fit, as produced by lm(), glm(), survival::coxph(), etc.
term	Term of interest
...	Term to condition on; ideally a vector. See examples.

Value

A data.frame suitable for use in ci_plot()

See Also

ci_plot()

Examples

fit <- lm(Ozone ~ Solar.R + Wind * Temp, airquality)
ci_int(fit, Wind, Temp=c(70, 80, 90))
ci_int(fit, Wind, Temp=c(70, 80, 90)) |> ci_plot()
airquality$Heat <- cut(airquality$Temp, 3, labels=c("Cool","Mild","Hot"))
fit <- lm(Ozone ~ Solar.R + Wind * Temp, airquality)
ci_int(fit, Wind, Temp) |> ci_plot()

# Survival
library(survival)
fit <- coxph(Surv(time, status) ~ karno * age, veteran)
ci_int(fit, karno, age) |> ci_plot()

# A transformation
fit <- lm(Ozone ~ Solar.R + log(Wind) * Temp, airquality)
ci_int(fit, log(Wind), Temp=c(70, 80, 90))
ci_int(fit, log(Wind), Temp=c(70, 80, 90)) |> ci_plot()
airquality$Heat <- cut(airquality$Temp, 3, labels=c("Cool","Mild","Hot"))
fit <- lm(Ozone ~ Solar.R + Wind * Heat, airquality)
ci_int(fit, Wind, Heat) |> ci_plot()

pbreheny/breheny documentation built on Feb. 18, 2025, 5:07 a.m.