Broom

broom

Convert statistical objects into tidy data frames.

Core Functions

library(broom)

# Fit a model
model %
  group_by(category) %>%
  nest() %>%
  mutate(
    model = map(data, ~ lm(y ~ x, data = .x)),
    tidied = map(model, tidy),
    glanced = map(model, glance)
  )

# Extract results
models %>%
  unnest(tidied)

models %>%
  unnest(glanced)

Augment for Diagnostics

model %
  filter(.cooksd > 4 / nrow(df))

Custom Tidiers

# For unsupported models, create custom tidier
tidy.my_model <- function(x, ...) {

  tibble(
    term = names(coef(x)),
    estimate = coef(x),
    std.error = sqrt(diag(vcov(x)))
  )
}

Supported Models

# broom supports 100+ model types including:
# - lm, glm, nls
# - survival (coxph, survfit)
# - lme4 (lmer, glmer)
# - mgcv (gam)
# - MASS (rlm, glm.nb)
# - stats (t.test, cor.test, chisq.test)
# - And many more...

# Check if model is supported
methods(tidy)
methods(glance)
methods(augment)

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: LeoLin990405
• Source: LeoLin990405/r-analytics-skill
• License: MIT

Install and usage instructions live in the source repository linked above.