Package Options

Quick read

Add these options to your .Rprofile for persistent use or add to the beginning of your table building script for ease of sharing.

options(sumExtras.auto_labels = TRUE)
options(sumExtras.prefer_dictionary = TRUE)

Auto-Labeling with sumExtras.auto_labels

If you keep a dictionary object in your environment (or your data already has label attributes), you can skip calling add_auto_labels() on every table. Set this once:

options(sumExtras.auto_labels = TRUE)

Now extras() handles labeling automatically:

# Define your dictionary once
dictionary <- tibble::tribble(
  ~variable,    ~description,
  "age",        "Age at Enrollment (years)",
  "marker",     "Marker Level (ng/mL)",
  "grade",      "Tumor Grade"
)

# Every extras() call picks it up
trial |>
  tbl_summary(by = trt) |>
  extras()

Put options(sumExtras.auto_labels = TRUE) in your .Rprofile to enable this for every session. Another option is to add it at the beginning of the script that creates your tables. If no dictionary is found and the data has no label attributes, extras() continues normally. If something goes wrong, it warns and moves on.

You can still call add_auto_labels() explicitly whenever you need per-table control.

Dictionary Priority with sumExtras.prefer_dictionary

By default, attribute labels (from attr(data$var, "label")) take priority over dictionary labels. If you maintain a centralized data dictionary and want it to win over attribute labels, set:

options(sumExtras.prefer_dictionary = TRUE)

This changes the label priority to:

1. Manual labels (from label = list(...) in tbl_summary()) – always win
2. Dictionary labels – from the dictionary data frame
3. Attribute labels – from attr(data$var, "label")

This is useful when your imported data has generic attribute labels (e.g., from {haven} or {labelled}) but your dictionary has the labels you actually want in publication tables.

Without the option (or with sumExtras.prefer_dictionary = FALSE), the default priority is manual > attributes > dictionary.

More Vignettes

• vignette("sumExtras-intro") – getting started with extras()
• vignette("labeling") – dictionary-based labeling and cross-package workflows
• vignette("themes") – JAMA compact themes for {gtsummary} and {gt} tables
• vignette("styling") – group headers and advanced formatting