broom (version 0.3.4)

coxph_tidiers: Tidiers for coxph object

Description

Tidy the coefficients of a Cox proportional hazards regression model, construct predictions, or summarize the entire model into a single row.

Usage

## S3 method for class 'coxph':
tidy(x, exponentiate = FALSE, conf.int = 0.95, ...)
## S3 method for class 'coxph':
augment(x, data = model.frame(x), newdata,
  type.predict = "lp", type.residuals = "martingale", ...)
## S3 method for class 'coxph':
glance(x, ...)

Arguments

x

"coxph" object

exponentiate

whether to report the estimate and confidence intervals on an exponential scale

conf.int

confidence level to be used for CI

...

Extra arguments, not used

data

original data for augment

newdata

new data on which to do predictions

type.predict

type of predicted value (see predict.coxph)

type.residuals

type of residuals (see residuals.coxph)

Value

• tidy returns a data.frame with one row for each term, with columns
• estimate
  estimate of slope
• std.error
  standard error of estimate
• statistic
  test statistic
• p.value
  p-value
• augment returns the original data.frame with additional columns added:
• .fitted
  predicted values
• .se.fit
  standard errors
• .resid
  residuals (not present if newdata is provided)
• glance returns a one-row data.frame with statistics calculated on the cox regression.

Details

When the modeling was performed with na.action = "na.omit" (as is the typical default), rows with NA in the initial data are omitted entirely from the augmented data frame. When the modeling was performed with na.action = "na.exclude", one should provide the original data as a second argument, at which point the augmented data will contain those rows (typically with NAs in place of the new columns). If the original data is not provided to augment and na.action = "na.exclude", a warning is raised and the incomplete rows are dropped.

See Also

na.action

Examples

if (require("survival", quietly = TRUE)) {
    cfit <- coxph(Surv(time, status) ~ age + sex, lung)

    tidy(cfit)
    tidy(cfit, exponentiate = TRUE)

    lp <- augment(cfit, lung)
    risks <- augment(cfit, lung, type.predict = "risk")
    expected <- augment(cfit, lung, type.predict = "expected")

    glance(cfit)

    # also works on clogit models
    resp <- levels(logan$occupation)
    n <- nrow(logan)
    indx <- rep(1:n, length(resp))
    logan2 <- data.frame(logan[indx,],
                         id = indx,
                         tocc = factor(rep(resp, each=n)))
    logan2$case <- (logan2$occupation == logan2$tocc)

    cl <- clogit(case ~ tocc + tocc:education + strata(id), logan2)
    tidy(cl)
    glance(cl)

    library(ggplot2)
    ggplot(lp, aes(age, .fitted, color = sex)) + geom_point()
    ggplot(risks, aes(age, .fitted, color = sex)) + geom_point()
    ggplot(expected, aes(time, .fitted, color = sex)) + geom_point()
}