regression_report.R in R

Fit a multiple regression and print a tidy coefficient table.

#!/usr/bin/env Rscript
# Fit a linear model on mtcars and print a tidy summary.

fit_model <- function(data) {
  lm(mpg ~ wt + hp + factor(cyl), data = data)
}

print_coefficients <- function(model) {
  coefs <- summary(model)$coefficients
  cat(sprintf("%-18s %9s %9s %9s\n",
              "term", "estimate", "std.error", "p.value"))
  for (term in rownames(coefs)) {
    cat(sprintf("%-18s %9.3f %9.3f %9.4f\n",
                term, coefs[term, 1], coefs[term, 2], coefs[term, 4]))
  }
}

model <- fit_model(mtcars)
print_coefficients(model)

cat("\nr.squared:    ", round(summary(model)$r.squared, 4), "\n")
cat("adj.r.squared:", round(summary(model)$adj.r.squared, 4), "\n")
cat("rmse:         ", round(sqrt(mean(resid(model)^2)), 4), "\n")
cat("aic:          ", round(AIC(model), 2), "\n")

How it works

1. The formula mixes numeric predictors with a factor.
2. summary(model)$coefficients drives a formatted table.
3. R-squared, RMSE, and AIC summarize the fit.

Keywords and builtins used here

• AIC
• cat
• factor
• fit_model
• for
• function
• in
• lm
• mean
• print_coefficients
• resid
• round
• rownames
• sprintf
• sqrt
• summary

The run, in numbers

Lines

24

Characters to type

752

Tokens

199

Three-star pace

105 tpm

At the three-star pace of 105 tokens a minute, this run takes about 114 seconds.

Type this snippet

Step 1 of 2 in Encore, step 10 of 11 in Statistics.

← Previous Next →