Skip to contents

Create a tidied tibble of regression results

Source: R/get_coefficients.R
get_coefficients.Rd

This function was created to produce results very similar to what you'll find at broom.helpers, with a few changes. Most notably, and the main reason for creating this function, you can standardize the regression coefficients by scaling and mean-centering the input data.

Usage

get_coefficients(
  model,
  conf.level = 0.95,
  standardize = FALSE,
  n.sd = 2,
  exponentiate = FALSE,
  add_ss = TRUE,
  add_labels = TRUE,
  add_n = FALSE,
  model_name = NULL
)

Arguments

model

A model object created using either lm or glm. Can also be piped into the function.

conf.level

A number between 0 and 1 that signifies the width of the desired confidence interval. Default is 0.95, which corresponds to a 95% confidence interval.

standardize

Logical. If TRUE, reports standardized regression coefficients by scaling and mean-centering input data. Default is FALSE.

n.sd

Logical. If standardize is TRUE, determines the number of standard deviations used to scale the data. Default is 2.

exponentiate

Logical. If TRUE, reports exponentiated coefficients with confidence intervals for exponential models like logit and Poisson models. This quantity is known as an odds ratio for binary outcomes and incidence rate ratio for count models. Default is FALSE.

add_ss

Logical. If TRUE, the default, a new column is created called ss that gives a "Yes" if the term is statistically significant and a "No" if the term is not statistically significant.

add_labels

Logical. If TRUE adds variable and value labels

add_n

Logical. If true adds the number of observations per variable

model_name

A character string that adds a new column titled model with the supplied character string as the rows. If NULL, the default, no column is created.

This is useful if you are comparing multiple models with similar variable and need to clarify which estimates are associated with which model.

Details

This function also takes advantage of tidy_add_reference_rows/, tidy_add_term_labels/, and tidy_add_n/ to allow you to include the reference row for each variable, the underlying variable and value labels, and the number of observations.