Get Started with ggforestplotR

library(ggforestplotR)
library(ggplot2)

ggforestplotR is built for coefficient-driven forest plots that stay inside a normal ggplot2 workflow.

Choose a workflow

Use the package in one of three ways:

  1. Start from a coefficient table and map the required columns directly.
  2. Start from a fitted model and let tidy_forest_model() or ggforestplot() call broom::tidy().
  3. Add add_forest_table() or add_split_table() after styling the main plot.

This article covers the basic entry points and the minimum data you need.

Start from a coefficient table

The simplest input is a data frame with a term, estimate, and confidence limits. If your columns use different names, map them explicitly.

basic_coefs <- data.frame(
  term = c("Age", "BMI", "Treatment"),
  estimate = c(0.10, -0.08, 0.34),
  conf.low = c(0.02, -0.16, 0.12),
  conf.high = c(0.18, 0.00, 0.56)
)

ggforestplot(basic_coefs)

Add grouped sections and row striping

Use grouping when you want related variables separated into labeled panels. Add striped_rows = TRUE to color alternating rows in the plot.

sectioned_coefs <- data.frame(
  term = c("Age", "BMI", "Smoking", "Stage II", "Stage III", "Nodes"),
  estimate = c(0.10, -0.08, 0.20, 0.34, 0.52, 0.28),
  conf.low = c(0.02, -0.16, 0.05, 0.12, 0.20, 0.06),
  conf.high = c(0.18, 0.00, 0.35, 0.56, 0.84, 0.50),
  section = c("Clinical", "Clinical", "Clinical", "Tumor", "Tumor", "Tumor")
)

ggforestplot(
  sectioned_coefs,
  grouping = "section",
  striped_rows = TRUE,
  stripe_fill = "grey94"
)

Add a summary table

Use add_forest_table() to add a summary table to your forest plot.

tabled_coefs <- data.frame(
  term = c("Age", "BMI", "Smoking", "Stage II", "Stage III"),
  estimate = c(0.12, -0.10, 0.18, 0.30, 0.46),
  conf.low = c(0.03, -0.18, 0.04, 0.10, 0.18),
  conf.high = c(0.21, -0.02, 0.32, 0.50, 0.74),
  sample_size = c(120, 115, 98, 87, 83)
)

ggforestplot(tabled_coefs, n = "sample_size", striped_rows = TRUE) +
  add_forest_table(
    position = "left",
    show_n = TRUE,
    estimate_label = "Beta"
  )

Add split summary tables

Use add_split_table() to create a more traditional looking forest plot, with summary data on either side of the plot.

ggforestplot(tabled_coefs, n = "sample_size", striped_rows = T) +
  add_split_table(
    left_columns = c("term", "n"),
    right_columns = c("estimate")
  )

Start from a fitted model

If broom is installed, ggforestplot() can work directly from a fitted model.

fit <- lm(mpg ~ wt + hp + qsec, data = mtcars)

ggforestplot(fit, sort_terms = "descending")