predict_CI.glm: Get predictions/CI from GLM

In Dasonk/glmExtras: glmExtras

Description

Gets predictions and confidence intervals on the response scale from a GLM object.

Usage

  ## S3 method for class 'glm'
 predict_CI(object, newdata = NULL,
    ci.level = 0.95, ...)

Arguments

object	The glm object
newdata	A data frame containing the columns necessary to make predictions from your glm object
ci.level	Numeric - A value between 0 and 1 that gives the confidence level for the confidence intervals you want created
...	Additional parameters to pass to predict

Examples

# Generating some fake data
n <- 30
x <- sort(rnorm(n))
link <- .2 - .8*x
p <- 1/(1 + exp(-link))
y <- rbinom(n, 1, p)
dat <- data.frame(x = x, y = y)

# Fit a logistic regression
o <- glm(y ~ x, data = dat, family = binomial)
# Get the predictions and confidence intervals
out <- predict_CI(o)
## Not run: 
# Make a plot
plot(x, y)
# true relationship
lines(x, p)
# estimated relationship
lines(x, out$prediction)
# confidence interval
lines(x, out$low.ci, col = "blue", lty = 2)
lines(x, out$up.ci, col = "blue", lty = 2)

## End(Not run)

Dasonk/glmExtras documentation built on May 6, 2019, 1:37 p.m.