Skip to contents

Repulsive textual annotations

Source: R/geom_dag.R
repel.Rd

These functions are minor modifications of those in the ggrepel package. geom_dag_text_repel() adds text directly to the plot. geom_dag_label_repel() draws a rectangle underneath the text, making it easier to read. The text labels repel away from each other and away from the data points. geom_dag_label_repel2() is a slightly stylized version of geom_dag_label_repel() that often looks better on DAGs. geom_dag_text_repel2() is a slightly stylized version of geom_dag_text_repel() that often looks better on DAGs.

Usage

geom_dag_text_repel(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  parse = FALSE,
  ...,
  node_size = NULL,
  n_edge_points = NULL,
  n_node_points = NULL,
  box.padding = 1.25,
  point.padding = 1,
  min.segment.length = 0.5,
  segment.color = "#666666",
  segment.alpha = 1,
  fontface = "bold",
  segment.size = 0.5,
  arrow = NULL,
  force = 1,
  force_pull = 1,
  max.time = 0.5,
  max.iter = 2000,
  max.overlaps = Inf,
  nudge_x = 0,
  nudge_y = 0,
  xlim = c(NA, NA),
  ylim = c(NA, NA),
  na.rm = FALSE,
  show.legend = NA,
  direction = c("both", "y", "x"),
  seed = NA,
  verbose = getOption("verbose", default = FALSE),
  inherit.aes = TRUE
)

geom_dag_label_repel(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  parse = FALSE,
  ...,
  node_size = NULL,
  n_edge_points = NULL,
  n_node_points = NULL,
  box.padding = grid::unit(1.25, "lines"),
  label.padding = grid::unit(0.25, "lines"),
  point.padding = grid::unit(1, "lines"),
  label.r = grid::unit(0.15, "lines"),
  label.size = 0.25,
  min.segment.length = 0.5,
  segment.color = "grey50",
  segment.alpha = 1,
  segment.size = 0.5,
  arrow = NULL,
  force = 1,
  force_pull = 1,
  max.time = 0.5,
  max.iter = 2000,
  max.overlaps = Inf,
  nudge_x = 0,
  nudge_y = 0,
  xlim = c(NA, NA),
  ylim = c(NA, NA),
  na.rm = FALSE,
  show.legend = NA,
  direction = c("both", "y", "x"),
  seed = NA,
  verbose = getOption("verbose", default = FALSE),
  inherit.aes = TRUE
)

geom_dag_label_repel2(
  mapping = NULL,
  data = NULL,
  box.padding = 2,
  max.overlaps = Inf,
  label.size = NA,
  linewidth = 0,
  ...
)

geom_dag_text_repel2(
  mapping = NULL,
  data = NULL,
  box.padding = 2,
  max.overlaps = Inf,
  ...
)

Arguments

mapping

Set of aesthetic mappings created by aes or aes_. If specified and inherit.aes = TRUE (the default), is combined with the default mapping at the top level of the plot. You only need to supply mapping if there isn't a mapping defined for the plot.

data

A data frame. If specified, overrides the default data frame defined at the top level of the plot.

stat

The statistical transformation to use on the data for this layer, as a string.

position

Position adjustment, either as a string, or the result of a call to a position adjustment function.

parse

If TRUE, the labels will be parsed into expressions and displayed as described in ?plotmath

...

other arguments passed on to layer. There are three types of arguments you can use here:

  • Aesthetics: to set an aesthetic to a fixed value, like colour = "red" or size = 3.

  • Other arguments to the layer, for example you override the default stat associated with the layer.

  • Other arguments passed on to the stat.

node_size

The size of the DAG nodes, used to compute the point.size aesthetic so that labels repel from the node boundary rather than the node center, and to size the skeleton discs described under Details. Defaults to NULL, which auto-discovers the size from a node layer (geom_dag_node() or geom_dag_point()) already added to the plot. Falls back to 16 if no node layer is found.

n_edge_points

Number of invisible points to interpolate along each edge. These "fake" points participate in ggrepel's repulsion calculation so that labels avoid overlapping edges. Defaults to NULL, which uses the StatNodesRepel default of 50. Set to 0 to disable edge-aware repulsion.

n_node_points

Target number of invisible points filling the disc that covers each node: a center point plus four concentric rings. Each ring holds at least six points, so every value from 1 to 16 gives the same 25 points per node and the parameter only starts to take effect above 16. Defaults to NULL, which uses the StatNodesRepel default of 12. Set to 0 to disable node skeleton repulsion.

box.padding

Amount of padding around bounding box, as unit or number. Defaults to 0.25. (Default unit is lines, but other units can be specified by passing unit(x, "units")).

point.padding

Amount of padding around labeled point, as unit or number. Defaults to 0. (Default unit is lines, but other units can be specified by passing unit(x, "units")).

min.segment.length

Skip drawing segments shorter than this, as unit or number. Defaults to 0.5. (Default unit is lines, but other units can be specified by passing unit(x, "units")).

segment.color, segment.size

See ggrepel::geom_text_repel()

segment.alpha

Transparency of the line segment. Set to NULL (default) to use ggrepel's default behavior, or provide a value between 0 and 1

fontface

A character vector. Default is "bold"

arrow

specification for arrow heads, as created by arrow

force

Force of repulsion between overlapping text labels. Defaults to 1.

force_pull

Force of attraction between a text label and its corresponding data point. Defaults to 1.

max.time

Maximum number of seconds to try to resolve overlaps. Defaults to 0.5.

max.iter

Maximum number of iterations to try to resolve overlaps. Defaults to 10000.

max.overlaps

Exclude text labels when they overlap too many other things. For each text label, we count how many other text labels or other data points it overlaps, and exclude the text label if it has too many overlaps. Defaults to 10.

nudge_x, nudge_y

Horizontal and vertical adjustments to nudge the starting position of each text label. The units for nudge_x and nudge_y are the same as for the data units on the x-axis and y-axis.

xlim, ylim

Limits for the x and y axes. Text labels will be constrained to these limits. By default, text labels are constrained to the entire plot area.

na.rm

If FALSE (the default), removes missing values with a warning. If TRUE silently removes missing values.

show.legend

logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes.

direction

"both", "x", or "y" – direction in which to adjust position of labels

seed

Random seed passed to set.seed. Defaults to NA, which means that set.seed will not be called.

verbose

If TRUE, some diagnostics of the repel algorithm are printed

inherit.aes

If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders.

label.padding

Amount of padding around label, as unit or number. Defaults to 0.25. (Default unit is lines, but other units can be specified by passing unit(x, "units")).

label.r

Radius of rounded corners, as unit or number. Defaults to 0.15. (Default unit is lines, but other units can be specified by passing unit(x, "units")).

label.size

Size of label border, in mm.

linewidth

Width of the label border in geom_dag_label_repel2(). Default is 0 (no border). Set to a positive value to show borders.

Details

These geoms are wrappers around ggrepel::geom_text_repel() and ggrepel::geom_label_repel() that use the custom StatNodesRepel for better handling of DAG data. All arguments available in ggrepel functions are supported.

Labels are kept off nodes and edges by two mechanisms. The point.size aesthetic, computed from node_size, is converted by ggrepel at draw time and so describes the same circle at every device size. The invisible points along edges and the disc filling each node, on the other hand, are placed in data units: the disc radius is node_size times the average spread of the nodes divided by 400 (the same divisor sizes the debug overlay described in ggdag_options_set()). Because the panel converts data units to millimetres at drawing time, that disc is congruent with the drawn node only on a panel about 180 mm wide. On a narrower device the disc is smaller than the node it stands for and a label may come to rest on the node; on a wider one it is larger and labels are pushed further away than they need to be. Set n_node_points = 0 to rely on point.size alone.

Points along an edge trace the path that edge is drawn along, including the arc of a bidirected edge and of geom_dag_edges_arc(). Edges drawn by geom_dag_edges_diagonal(), geom_dag_edges_fan(), and the ggarrow engine are traced along the straight line between their nodes.

Additional segment parameters can be passed through ..., including:

  • segment.linetype: Line style

  • segment.alpha: Line transparency

  • segment.curvature: Curve amount

  • segment.angle: Curve angle

  • segment.ncp: Number of control points

  • segment.shape: Control point position

  • segment.square: Square formation control points

  • segment.squareShape: Square formation shape

  • segment.inflect: Add inflection point

  • segment.debug: Show debug information

You can also pass point.size and point.colour through ....

Examples

library(ggplot2)
g <- dagify(
  m ~ x + y,
  y ~ x,
  exposure = "x",
  outcome = "y",
  latent = "m",
  labels = c("x" = "Exposure", "y" = "Outcome", "m" = "Collider")
)

g |>
  tidy_dagitty() |>
  ggplot(aes_dag()) +
  geom_dag_edges() +
  geom_dag_point() +
  geom_dag_text_repel(aes(label = name), show.legend = FALSE) +
  theme_dag()


# Use nudge_x and nudge_y to push labels away from nodes
g |>
  tidy_dagitty() |>
  ggplot(aes_dag()) +
  geom_dag_edges() +
  geom_dag_point() +
  geom_dag_text_repel(
    aes(label = name),
    nudge_x = 0.1,
    nudge_y = 0.1
  ) +
  theme_dag()


# Use position_nudge_repel for the same effect
g |>
  tidy_dagitty() |>
  ggplot(aes_dag()) +
  geom_dag_edges() +
  geom_dag_point() +
  geom_dag_text_repel(
    aes(label = name),
    position = ggrepel::position_nudge_repel(x = 0.1, y = 0.1)
  ) +
  theme_dag()


g |>
  tidy_dagitty() |>
  dag_label(labels = c(
    "x" = "This is the exposure",
    "y" = "Here's the outcome",
    "m" = "Here is where they collide"
  )) |>
  ggplot(aes_dag()) +
  geom_dag_edges() +
  geom_dag_point() +
  geom_dag_text() +
  geom_dag_label_repel(
    aes(label = label, fill = label),
    col = "white",
    show.legend = FALSE
  ) +
  theme_dag()


# Use directional repulsion
g |>
  tidy_dagitty() |>
  ggplot(aes_dag()) +
  geom_dag_edges() +
  geom_dag_point() +
  geom_dag_text_repel(
    aes(label = name),
    direction = "y",
    seed = 1234
  ) +
  theme_dag()


# Customize segment appearance
g |>
  tidy_dagitty() |>
  ggplot(aes_dag()) +
  geom_dag_edges() +
  geom_dag_point() +
  geom_dag_text_repel(
    aes(label = name),
    segment.linetype = 2,
    segment.alpha = 0.5,
    segment.curvature = -0.3
  ) +
  theme_dag()