In jeksterslabds/jeksterslabRlinreg: Jekster's Lab Linear Regression Utilities

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.width = 5,
  fig.height = 5,
  out.width = "100%"
)

jeksterslabRlinreg is a collection of functions that I find useful in studying linear regression concepts and methods.

Installation

You can install the released version of jeksterslabRlinreg from GitHub with:

library(devtools)
install_github("jeksterslabds/jeksterslabRlinreg")

Example

Data

In this hypothetical example, we are interested in the association between wages and education. The regressor variable is years of education. The regressand variable is hourly wage in US dollars.

wages <- jeksterslabRdatarepo::wages
x <- wages$education
y <- wages$wages
n <- length(x)
obj <- lm(y ~ x)
beta <- unname(coef(obj))
# covariance structure
beta1 <- beta[1]
beta2 <- beta[2]
sigma2x <- var(x)
sigma2y <- var(y)
# sigma^2 has some discrepancy
# sigma2epsilon <- summary(obj)$sigma^2
sigma2epsilon <- sigma2y - (beta2^2 * sigma2x)
# mean structure
mux <- mean(x)

x <- as.integer(rnorm(n = n, mean = mux, sd = sqrt(sigma2x)))
epsilon <- rnorm(n = n, mean = 0, sd = sqrt(sigma2epsilon))
y <- beta1 + beta2 * x + epsilon
X <- cbind(
  constant = 1,
  education = x
)
y <- matrix(
  data = y,
  ncol = 1
)
colnames(y) <- "wages"
head(X)
head(y)

jeksterslabRlinreg::linreg()

The jeksterslabRlinreg::linreg() function fits a linear regression model using X and y. In this example, X consists of a column of constants and years of education and y consists of hourly wages in US dollars.

The output includes the following:

• Model assessment
• ANOVA table
• Table of regression coefficients with the following columns
• Regression coefficients
• Standard errors
• $t$ statistic
• $p$ value
• Standardized coefficients
• Confidence intervals (0.05, 0.5, 2.5, 97.5, 99.5, 99.95)
• Means and standard deviations
• Scatterplot matrix
• Residual plots

jeksterslabRlinreg::linreg(
  X = X,
  y = y
)

jeksterslabds/jeksterslabRlinreg documentation built on Jan. 7, 2021, 8:30 a.m.