Modify operated Context in quad_layout()

The with_quad() function modifies the application context of elements in ggheatmap()/quad_layout(). It controls how objects like themes, scales, or other plot modifications apply to specific annotation stacks or the main plot without altering the currently active layout or plot.

Usage

with_quad(x, position = waiver(), main = NULL)

Arguments

x: An object which can be added to the ggplot, including schemes. See scheme_align(), scheme_data(), and scheme_theme()
position: A string specifying one or more positions- "t", "l", "b", and "r"- to indicate the annotation stack context for x. If NULL, will change the operated context to the quad_layout() itself. For default behaivours, see details section.
main: A single boolean value indicating whether x should apply to the main plot, used only when position is not NULL. By default, if position is waiver() and the active context of quad_layout() is an annotation stack or the active context of stack_layout() is itself, main will be set to TRUE; otherwise, it defaults to FALSE.

Value

The original object with an added attribute that sets the specified context.

Details

Default Behavior when adding object wrapped with with_quad():

For quad_layout() object:

When ggheatmap()/quad_layout() has no active annotation stack, objects added via + or - operate normally without with_quad().
When the active annotation stack is set, with_quad() ensures the applied object also modifies:
- The main plot (by default).
- Opposite annotation stacks when using -.

For stack_layout() object:

When the active layout is the stack_layout() itself:
- - operator will apply changes to all plots along the stack_layout(), which means if the stack layout is in horizontal, - operator will also add the element to the left and right annotation, if the stack layout is in vertical, - operator will also add element to the top and bottom annotation.
- + operator won't do anything special.
When the active layout is the nested ggheatmap()/quad_layout(), the +/- operator applies the elements to this nested layout, following the same principles as for ggheatmap()/quad_layout().

Examples

set.seed(123)
small_mat <- matrix(rnorm(56), nrow = 7)

# By wrapping object with `with_quad()`, the `+` operator will apply the
# object not only to the active plot in the annotation stack, but also to
# the main plot unless specified by `main` argument otherwise.
ggheatmap(small_mat) +
    # initialize the left annotation
    anno_left(size = 0.2) +
    align_dendro() +
    # apply the object not only to the active plot in the annotation stack,
    # but also to the main plot
    with_quad(theme(plot.background = element_rect(fill = "red")))
#> → heatmap built with `geom_tile()`


# the `-` operator will apply changes not only to the active annotation
# stack but also to the opposite one (i.e., bottom if top is active, and
# vice versa). The same principle applies to the left and right annotation.
ggheatmap(small_mat) +
    anno_left(size = 0.2) +
    align_dendro(aes(color = branch), k = 3L) +
    # Change the active layout to the left annotation
    anno_top(size = 0.2) +
    align_dendro(aes(color = branch), k = 3L) +
    anno_bottom(size = 0.2) +
    align_dendro(aes(color = branch), k = 3L) -
    # Modify the color scale of all plots in the bottom and the opposite
    # annotation, in this way, the `main` argument by default would be `TRUE`
    with_quad(scale_color_brewer(palette = "Dark2", name = "Top and bottom"))
#> → heatmap built with `geom_tile()`


# When the `position` argument is manually set, the
# default value of the `main` argument will be `FALSE`.
ggheatmap(small_mat) +
    anno_left(size = 0.2) +
    align_dendro(aes(color = branch), k = 3L) +
    anno_top(size = 0.2) +
    align_dendro(aes(color = branch), k = 3L) +
    anno_bottom(size = 0.2) +
    align_dendro(aes(color = branch), k = 3L) -
    # Modify the background of all plots in the left and top annotation
    with_quad(theme(plot.background = element_rect(fill = "red")), "tl")
#> → heatmap built with `geom_tile()`