Side-by-Side Forest Plots: Univariate vs Multivariable — plot_reg

Creates two aligned forest plots (univariate and multivariable) from `gtsummary`-style objects returned by `gtregression` functions (e.g., `uni_reg()`, `multi_reg()`).

Usage

plot_reg_combine(
  tbl_uni,
  tbl_multi,
  title_uni = NULL,
  title_multi = NULL,
  ref_line = NULL,
  order_y = NULL,
  log_x = FALSE,
  point_color = "#1F77B4",
  errorbar_color = "#4C4C4C",
  base_size = 14,
  show_ref = TRUE,
  sig_color = NULL,
  sig_errorbar_color = NULL,
  xlim_uni = NULL,
  breaks_uni = NULL,
  xlim_multi = NULL,
  breaks_multi = NULL,
  alpha = 0.05
)

Arguments

tbl_uni: Univariate `gtsummary`-like table.
tbl_multi: Multivariable `gtsummary`-like table.
title_uni, title_multi: Optional panel titles.
ref_line: Optional numeric reference line (defaults to 0 for linear, 1 otherwise, inferred per panel).
order_y: Optional character vector to customize header ordering.
log_x: Logical. If `TRUE`, use log x-axis (ignored for linear models).
point_color, errorbar_color: Base colors for non-significant rows.
base_size: Base font size for `theme_minimal()`.
show_ref: Logical; if `TRUE`, include and tag reference levels `(Ref.)`.
sig_color, sig_errorbar_color: Optional colors for significant rows; if `NULL`, they reuse the base colors.
xlim_uni, breaks_uni: Optional x-limits and breaks for the univariate panel.
xlim_multi, breaks_multi: Optional x-limits and breaks for the multivariable panel.
alpha: Significance level for linear models when `p.value` is available.

Value

A `patchwork` object with two `ggplot2` panels.

Details

The y-axis rows are aligned by a unique `(variable, level)` key so each estimate appears exactly once per panel. Label styling is plain text by default (CRAN-safe). To render bold headers / grey refs in vignettes, pair

Examples

# \donttest{
if (requireNamespace("mlbench", quietly = TRUE) &&
    requireNamespace("gtregression", quietly = TRUE)) {
  data("PimaIndiansDiabetes2", package = "mlbench")
  d <- PimaIndiansDiabetes2
  d$diabetes <- ifelse(d$diabetes == "pos", 1, 0)

  tbl_u <- gtregression::uni_reg(d, outcome = "diabetes",
                                 exposures = c("age","glucose"), approach = "logit")
  tbl_m <- gtregression::multi_reg(d, outcome = "diabetes",
                                   exposures = c("age","glucose"), approach = "logit")
  plot_reg_combine(tbl_u, tbl_m,
                   title_uni = "Univariate", title_multi = "Adjusted")
}

# }