simulate: Simulate a data frame for multiple regression.

In jbryer/VisualStats: Visualizations for Statistical Tests

Simulate a data frame for multiple regression.

Description

Simulate a data frame for multiple regression.

Usage

simulate(n = 100, beta = c(5, 3, -2), r_squared = 0.8)

Arguments

n	number of observations.
beta	beta coefficients. Position one is the intercept, the rest are predictors.
r_squared	desired R-squared value.

Value

a data.frame.

Examples

get_r_squared <- function(desired) {
    model <- lm(y ~ x1 + x2, data = simulate(r_squared = desired, n = 100))
    return(summary(model)$r.squared)
}
df <- data.frame(desired_r_squared = seq(from = 0.05, to = 0.95, by = 0.05))
df$actual_r_squared <- sapply(df$desired_r_squared, FUN = get_r_squared)
plot(df)
abline(a=0, b=1, col="red", lty=2)

jbryer/VisualStats documentation built on Feb. 27, 2025, 6:19 p.m.