R: Retrieve labels of model terms from regression models

get_term_labels {sjlabelled}

R Documentation

Retrieve labels of model terms from regression models

Description

This function retrieves variable labels from model terms. In case of categorical variables, where one variable has multiple dummies, variable name and category value is returned.

Usage

get_term_labels(models, mark.cat = FALSE, case = NULL, ...)

get_dv_labels(models, case = NULL, ...)

Arguments

`models`	One or more fitted regression models. May also be glm's or mixed models. Any model with `model.frame` method and which is supported by broom's `tidy` function should work.
`mark.cat`	Logical, if `TRUE`, the returned vector has an attribute with logical values, which indicate whether a label indicates the value from a factor category (attribute value is `TRUE`) or a term's variable labels (attribute value is `FALSE`).
`case`	Desired target case. Labels will automatically converted into the specified character case. See `to_any_case` for more details on this argument.
`...`	Further arguments passed down to `to_any_case`, like `preprocess` or `postprocess`.

Details

Typically, the variable labels from model terms are returned. However, for categorical terms that have estimates for each category, the value labels are returned as well. As the return value is a named vector, you can easily use it with ggplot2's scale_*() functions to annotate plots.

Value

For get_term_labels(), a (named) character vector with variable labels of all model terms, which can be used, for instance, as axis labels to annotate plots.

For get_dv_labels(), a character vector with variable labels from all dependent variables of models.

Examples

# use data set with labelled data
data(efc)

fit <- lm(barthtot ~ c160age + c12hour + c161sex + c172code, data = efc)
get_term_labels(fit)

# make "education" categorical
library(sjmisc)
efc$c172code <- to_factor(efc$c172code)
fit <- lm(barthtot ~ c160age + c12hour + c161sex + c172code, data = efc)
get_term_labels(fit)

# get label of dv
get_dv_labels(fit)

# create "labelled" plot
library(ggplot2)
library(broom)
ggplot(tidy(fit)[-1, ], aes(x = term, y = estimate)) +
  geom_point() +
  coord_flip() +
  scale_x_discrete(labels = get_term_labels(fit))

[Package sjlabelled version 1.0.8 Index]