ggdag

ggdag is designed to work with the dagitty package. It provides a tidyverse-friendly interface to create, analyze, and plot causal directed acyclic graphs (DAGs).

You can use ggdag in two primary ways:

1. Tidying existing dagitty objects: Use tidy_dagitty(dag) to convert a dagitty object into a tidy data frame suitable for ggplot2.
2. Creating DAGs with R-like syntax: Use the dagify() function to define causal relationships using formula syntax, which is more intuitive for R users than the dagitty string syntax.

To retrieve the original DAG object from a tidied object, use pull_dag(). To retrieve the underlying data frame, use pull_dag_data().

library(ggdag)
library(ggplot2)

# Example: Creating a DAG using dagify syntax
tidy_ggdag <- dagify(
  y ~ x + z2 + w2 + w1,
  x ~ z1 + w1 + w2,
  z1 ~ w1 + v,
  z2 ~ w2 + v,
  w1 ~ ~w2, # bidirected path
  exposure = "x",
  outcome = "y",
  coords = time_ordered_coords()
) |> 
  tidy_dagitty()

Once a DAG is tidied, you can plot it using ggdag() and standard ggplot2 layers.

Basic Plotting:

ggdag(tidy_ggdag) + theme_dag()

Plotting Adjustment Sets: Use ggdag_adjustment_set() to visualize adjustment sets.

ggdag_adjustment_set(tidy_ggdag, node_size = 14) + theme(legend.position = "bottom")

Advanced Customization: You can use specific ggdag geoms and scales for fine-grained control over edges, nodes, and text:

• geom_dag_edges(): Draws edges (supports end_cap).
• geom_dag_collider_edges(): Draws collider edges.
• geom_dag_point(): Draws nodes.
• geom_dag_text(): Draws node labels.
• theme_dag(): A specialized theme for DAGs.
• scale_adjusted(): Scales for adjusted/unadjusted status.
• node_dconnected(): A function to identify d-connected nodes for specific paths.

dagify(m ~ x + y) |
  tidy_dagitty() |
  node_dconnected("x", "y", controlling_for = "m") |
  ggplot(aes(
    x = x,
    y = y,
    xend = xend,
    yend = yend,
    shape = adjusted,
    col = d_relationship
  )) +
  geom_dag_edges(end_cap = ggraph::circle(10, "mm")) +
  geom_dag_collider_edges() +
  geom_dag_point() +
  geom_dag_text(col = "white") +
  theme_dag() +
  scale_adjusted(include_color = FALSE) +
  expand_plot(expand_y = expansion(c(0.2, 0.2))) +
  scale_color_viridis_d(
    name = "d-relationship",
    na.value = "grey85",
    begin = .35
  )