loon.ggplot: A Grammar of Interactive Graphics

Description

Provides a bridge between the 'loon' and 'ggplot2' packages. Extends the grammar of ggplot to add clauses to create interactive 'loon' plots. Existing ggplot(s) can be turned into interactive 'loon' plots and 'loon' plots into static ggplot(s); the function 'loon.ggplot()' is the bridge from one plot structure to the other.

Author(s)

Maintainer: Zehao Xu z267xu@gmail.com

Authors:

• R. Wayne Oldford rwoldford@uwaterloo.ca

See Also

Useful links:

• Report bugs at https://github.com/great-northern-diver/loon.ggplot/issues

Transform the x, y positions from a Cartesian coordinate to a polar coordinate

Description

Used in the 'loonLayer' construction to access the x, y positions embedded in the polar coordinate system.

Usage

Cartesianxy2Polarxy(layerGeom, coordinates, data, ggplotPanelParams, ...)

Arguments

Examples

p <- ggplot(mtcars, aes(wt, mpg)) +
       geom_point() +
       coord_polar()

layerGeom <- p$layers[[1L]]$geom
coordinates <- p$coordinates
build <- ggplot_build(p)
data <- build$data[[1L]]
ggplotPanelParams <- build$layout$panel_params[[1L]]

polarXY <- Cartesianxy2Polarxy(layerGeom, coordinates, data, ggplotPanelParams)
plot(polarXY$x, polarXY$y)

Modify the active component

Description

Set active and/or activeGeomLayers

Usage

active(active = NULL, activeGeomLayers = NULL)

Arguments

Value

a ggproto object

See Also

linking, selection, zoom, hover, interactivity

Examples

if(interactive()) {

  # set active layer
  l_ggplot(mtcars, aes(mpg, wt, shape = factor(cyl))) +
    geom_point(colour = "black", size = 4.5) +
    geom_point(colour = "pink", size = 4) +
    geom_point(aes(shape = factor(cyl))) +
    # only show manual transmission cars
    # in the second interactive layer
    active(active = mtcars$am == 1,
           activeGeomLayers = 2)

  # Then, click the `reactivate` button on loon inspector
  # to display all interactive points
}

get locations for ggmatrix

Description

For the target compound loon plot, determines location in ggmatrix

Usage

g_getLocations(target)

## Default S3 method:
g_getLocations(target)

## S3 method for class 'l_pairs'
g_getLocations(target)

Arguments

Value

a list of an appropriate subset of the named location arguments 'c("ncol", "nrow", "layout_matrix", "heights", "widths")'. layout_matrix is an nrow by ncol matrix whose entries identify the location of each plot in g_getPlots() by their index.

See Also

l_getLocations, g_getPlots

get ggplots

Description

For the target compound loon plot, determines all the ggplots based on the compound loon plot.

Usage

g_getPlots(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE
)

## Default S3 method:
g_getPlots(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE
)

## S3 method for class 'l_pairs'
g_getPlots(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE
)

Arguments

Value

a list of ggplots.

See Also

l_getPlots, g_getLocations

geom_imageGlyph (deprecated)

Description

use geom_image_glyph, rather than geom_imageGlyph.

use geom_polygon_glyph, rather than geom_polygonGlyph.

Usage

geom_imageGlyph(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  ...,
  images,
  width = 4,
  height = 3,
  na.rm = FALSE,
  show.legend = NA,
  inherit.aes = TRUE
)

geom_polygonGlyph(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  ...,
  polygon_x,
  polygon_y,
  showArea = TRUE,
  linewidth = 1,
  na.rm = FALSE,
  show.legend = NA,
  inherit.aes = TRUE
)

Arguments

geom_pointrangeGlyph (deprecated)

Description

use geom_pointrange, rather than geom_pointrangeGlyph.

Usage

geom_pointrangeGlyph(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  ...,
  ymin,
  ymax,
  showArea = TRUE,
  linewidth = 1,
  na.rm = FALSE,
  show.legend = NA,
  inherit.aes = TRUE
)

Arguments

geom_serialAxesGlyph (deprecated)

Description

use geom_serialaxes_glyph, rather than geom_serialAxesGlyph.

Usage

geom_serialAxesGlyph(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  ...,
  serialAxesData,
  sequence = NULL,
  linewidth = 1,
  scaling = c("variable", "data", "observation", "none"),
  axesLayout = c("parallel", "radial"),
  showAxes = FALSE,
  showArea = FALSE,
  showEnclosing = FALSE,
  axesColor = "black",
  bboxColor = "black",
  na.rm = FALSE,
  show.legend = NA,
  inherit.aes = TRUE
)

Arguments

geom_textGlyph (deprecated)

Description

use geom_text, rather than geom_textGlyph.

Usage

geom_textGlyph(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  ...,
  text,
  na.rm = FALSE,
  show.legend = NA,
  inherit.aes = TRUE
)

Arguments

active geom layers

Description

'get_activeGeomLayers' will return the geom layer index which can be active

Usage

get_activeGeomLayers(ggObj)

Arguments

Details

'ggplot2loon' has an argument called 'activeGeomLayers'. It is a vector to determine which geom layers can be active. The default setting is 'integer(0)', however, 'ggplot2loon' will automatically search the first 'geom_histogram' or 'geom_point' layer to make it active. 'get_activeGeomLayers' is more like a guidance and give us a hint which one can be set as active.

Value

a numerical vector of indicies (which layer can be interactive)

See Also

ggplot2loon

Examples

if(interactive()) {

  df <- data.frame(x = 1:3, y = 1:3, colour = c(1,3,5))
  xgrid <- with(df, seq(min(x), max(x), length = 50))
  interp <- data.frame(
    x = xgrid,
    y = approx(df$x, df$y, xout = xgrid)$y,
    colour = approx(df$x, df$colour, xout = xgrid)$y
  )
  p1 <- ggplot(data = df, aes(x, y, colour = colour)) +
    geom_line(interp, mapping = aes(x, y, colour = colour), size = 2) +
    geom_point(size = 5)
  agL <- get_activeGeomLayers(p1)
  ggplot2loon(p1, activeGeomLayers = agL)

  p2 <- ggplot(economics) +
    geom_rect(
      aes(xmin = start, xmax = end, fill = party),
      ymin = -Inf, ymax = Inf, alpha = 0.2,
      data = presidential
    ) +
    geom_text(
      aes(x = start, y = 2500, label = name),data = presidential,
      size = 3, vjust = 0, hjust = 0, nudge_x = 50
    ) +
    geom_line(aes(date, unemploy)) +
    scale_fill_manual(values = c("blue", "red"))
  # none can be interactive
  agL <- get_activeGeomLayers(p2)
  #transparency is not allowed in tcltk
  ggplot2loon(p2, ggGuides = TRUE, activeGeomLayers = agL)

}

scale data

Description

It is mainly used in serial axes

Usage

get_scaledData(
  data,
  sequence = NULL,
  scaling = c("variable", "data", "observation", "none"),
  displayOrder = NULL,
  keep = FALSE,
  as.data.frame = FALSE
)

Arguments

ggplot serialaxes (deprecated)

Description

For ggSerialAxes, use coord_serialaxes.

Usage

ggSerialAxes(
  ggObj,
  data = NULL,
  axesLabels = NULL,
  showAxes = TRUE,
  showAxesLabels = TRUE,
  scaling = c("variable", "observation", "data", "none"),
  layout = c("parallel", "radial"),
  displayOrder = NULL,
  title = "",
  showLabels = TRUE,
  color = NULL,
  size = NULL,
  showGuides = TRUE,
  showArea = FALSE
)

Arguments

Value

a ggplot object

Examples

## Not run: 
# Blank plot
p <- ggplot(data = mtcars, mapping = aes(colour = factor(cyl)))
# Add serial axes (returns a ggplot object)
g <- ggSerialAxes(p)
g
# An eulerian path of iris variables
# ordSeq <- PairViz::eulerian(4)
ordSeq <- c(1, 2, 3, 1, 4, 2, 3, 4)
ggSerialAxes(
       ggObj = ggplot(data = iris, mapping = aes(colour = Species)),
       axesLabels = colnames(iris)[ordSeq],
       layout = "radial"
)

## End(Not run)

Pipe ggplot object

Description

Pack a ggplot object forward to ggplot2loon expressions via a pipe-operator "%>%".

Usage

gg_pipe(data, ggObj)

Arguments

Details

When "+" and "%>%" both appear in pipe operations, "%>%" takes the priority of "+",e.g:

mtcars %>% ggplot(aes(mpg, wt, colour = cyl)) + geom_point() %>% ggplot2loon(),

error would occur. The reason is

geom_point() %>% ggplot2loon()

would run before

ggplot(aes(mpg, wt, colour = cyl)) + geom_point().

Hence, we need a function gg_pipe() to pack the ggplot object and force operations happen in order.

Value

a ggplot evaluate object

Examples

if(requireNamespace("magrittr") && interactive()) {
## Not run: 
 # Error
 g <- mtcars %>%
    ggplot(aes(mpg, wt, colour = cyl)) +
    geom_point() %>%
    ggplot2loon()

## End(Not run)
 g <- mtcars %>%
   gg_pipe(
     ggplot(aes(mpg, wt, colour = cyl)) + geom_point()
   ) %>%
   ggplot2loon()
}

ggplot to loon

Description

Create an interactive 'loon' widget from a ggplot object

Usage

ggplot2loon(
  ggObj,
  ...,
  activeGeomLayers = integer(0),
  layerId = NULL,
  scaleToFun = NULL,
  ggGuides = FALSE,
  parent = NULL,
  pack = TRUE,
  exteriorLabelProportion = 1/5,
  canvasHeight = 700,
  canvasWidth = 850,
  tkLabels = NULL
)

Arguments

Value

a loon single widget or a compound object

Examples

if(interactive()) {
p <- ggplot(mtcars, aes(wt, mpg)) + geom_point()
g <- ggplot2loon(p)

p1 <- ggplot(mtcars) +
        geom_point(aes(x = wt, y = mpg,
                       colour = factor(gear))) +
        facet_wrap(~am)
g1 <- ggplot2loon(p1)

df <- data.frame(
  x = rnorm(120, c(0, 2, 4)),
  y = rnorm(120, c(1, 2, 1)),
  z = letters[1:3]
)
df2 <- dplyr::select(df, -z)
scatterplots <- ggplot(df, aes(x, y)) +
  geom_point(data = df2, colour = "grey70") +
  geom_point(aes(colour = z)) +
  facet_wrap(~z)

# The first point layer is set as the model layer
suppressWarnings(
  lp_scatterplots_active1 <- ggplot2loon(scatterplots,
                               activeGeomLayers = 1,
                               linkingGroup = "test")
)
# Here, the gray points are interactive (not the colourful ones)

# The second point layer is set as the model layer
lp_scatterplots_active2 <- ggplot2loon(scatterplots,
                                       activeGeomLayers = 2)
# Here, the colourful points are interactive

# Both point layers could be interactive
suppressWarnings(
 lp_scatterplots_active12 <- ggplot2loon(scatterplots,
                                         activeGeomLayers = c(1,2))
)
# Here, all points are interactive

########### ggmatrix to loon ###########
if(requireNamespace("GGally")) {
pm <- GGally::ggpairs(iris, column = 1:4,
                      ggplot2::aes(colour=Species))
lg <- ggplot2loon(pm)
}

########### patchwork to loon ###########
if(requireNamespace("patchwork")) {
p1 <- ggplot(mtcars) +
   geom_point(aes(mpg, disp))
p2 <- ggplot(mtcars) +
   geom_boxplot(aes(gear, disp, group = gear))
# place two plots side by side
patchwork <- p1 + p2
ggplot2loon(patchwork)
# See vignette `ggplots --> loon plots` for more details
}

}

Modify the hover component

Description

Provides a pop up display as the mouse hovers over a plot element in the interactive plot.

Usage

hover(itemLabel = NULL, showItemLabels = NULL)

Arguments

Value

a ggproto object

See Also

active, linking, zoom, selection, interactivity

Examples

if(interactive()) {

  l_ggplot(mpg, mapping = aes(x = displ, y = cty)) +
    geom_point(size = 4) +
    # push the states of scatter plot to the histogram
    hover(itemLabel =
       with(mpg,
            paste0("model: ", manufacturer, " ", model, "\n",
                   "year: ", year, "\n",
                   "drive way: ", drv, "\n",
                   "fuel type: ", fl)
       ),
       showItemLabels = TRUE
    )
    # hover the mouse on top of any point to query
}

Modify the interactivity component

Description

Set interactive components (e.g. linking, selection, etc)

Usage

interactivity(
  linkingGroup = NULL,
  linkingKey = NULL,
  linkedStates = NULL,
  sync = NULL,
  active = NULL,
  activeGeomLayers = NULL,
  selected = NULL,
  selectBy = NULL,
  selectionLogic = NULL,
  layerId = NULL,
  scaleToFun = NULL,
  itemLabel = NULL,
  showItemLabels = NULL,
  ...
)

Arguments

Details

In interactive graphics, there are several fundamental infrastructures, such as querying, linking and selection. Component interactivity is used to set these features.

Value

a ggproto object

Examples

if(interactive()) {
  # Modify the 'linkingGroup' and 'origin' of a hist object
  l_ggplot(mtcars, mapping = aes(x = wt)) +
    geom_histogram() +
    interactivity(linkingGroup = "mt", origin = 2)

  # linking with the histogram
  l_ggplot(mtcars, mapping = aes(x = wt, y = hp)) +
    geom_point(size = 4) +
    interactivity(linkingGroup = "mt") +
    facet_wrap(~cyl)

  p <- ggplot(economics_long, aes(value)) +
         facet_wrap(~variable, scales = 'free_x') +
         geom_histogram()
  # `p` is a ggplot object
  p
  # turn static `ggplot` to interactive `loon`
  p + interactivity()
}

Is polar coordinate system?

Description

Determine whether the ggplot object has polar coordinate system

Usage

is.CoordPolar(coord)

Arguments

Reports whether x is a l_ggplot object

Description

Reports whether x is a l_ggplot object

Usage

is.l_ggplot(x)

Arguments

itemLabel (deprecated)

Description

use hover, rather than itemLabel.

Usage

itemLabel(itemLabel = NULL, showItemLabels = NULL)

Arguments

Return the subtitles

Description

Return the subtitles

Usage

l_getSubtitles(target)

## S3 method for class 'l_facet_ggplot'
l_getSubtitles(target)

## S3 method for class 'l_facet_wrap'
l_getSubtitles(target)

## S3 method for class 'l_facet_grid'
l_getSubtitles(target)

Arguments

Value

A list of labels, i.e. subtitles, labels, title, etc

Examples

if(interactive()) {
p <- ggplot(mpg, aes(displ, hwy)) +
  geom_point() +
  facet_wrap(vars(class))
lp <- loon.ggplot(p)
l_getSubtitles(lp)
}

Automatically create a loon widget

Description

Create a loon widget with ggplot syntax

Usage

l_ggplot(data = NULL, mapping = aes(), ..., environment = parent.frame())

Arguments

Details

function l_ggplot() wraps function ggplot() with assigning an additional class "l_ggplot" to the output. The returned object is called an l_ggplot object. To draw a ggplot object, S3 method print.ggplot will be rendered so that a static graphic is displayed. While, for an l_ggplot() object, S3 method print.l_ggplot will be rendered which will return an interactive loon widget.

Value

It will return an l_ggplot object with class c("l_ggplot", "gg", "ggplot"). Then print a loon plot automatically.

See Also

ggplot, ggplot2loon, print.l_ggplot

loon.ggplot

Examples

if(interactive()) {
p <- l_ggplot(mpg, aes(displ, cty)) +
    geom_point(
      size = 4,
      mapping = aes(color = factor(cyl))
    )
# p is an `l_ggplot` object, `print.l_ggplot(p)` will be called automatically.
# Then, at printing time, an `l_ggplot` object will be transformed to a `loon` widget
p

## Not run: 
# Assign a widget from current path
# suppose the path of `p` is '.l0.ggplot'
q <- l_getFromPath('.l0.ggplot')
# q is a `loon` widget
q

## End(Not run)

# An alternative way to return a real loon widget from `p` (a `l_ggplot` object)
# is to call the function `loon.ggplot()`.
q <- loon.ggplot(p)
q

# pipe more components
p +
  facet_grid(rows = vars(drv)) +
  linking(linkingGroup = "mpg") +
  ggtitle("displ versus cty")
# a linked bar plot
l_hist(mpg$class, linkingGroup = "mpg")

# a 3D object
# press the button key `R` to rotate the plot
l_ggplot(mtcars,
         mapping = aes(x = wt, y = hp, z = drat)) +
   geom_point(size = 4) +
   scale_multi()
}

layout matrix

Description

return the layout matrix of a list of loon plots

Usage

layout_coords(target)

Arguments

Value

a layout coordinate matrix

lggplot (deprecated)

Description

use l_ggplot, rather than lggplot.

Usage

lggplot(data = NULL, mapping = aes(), ..., environment = parent.frame())

Arguments

Modify the linking component

Description

A group-key-state linking model is used to link plots in loon. This allows changes in one plot to propogate to all plots in the same linkingGroup and enables interactive features like brushing. Elements to be matched between plots are identified by linkingKey; within each plot, the key for each element (e.g., case, observation) is unique. The linkedStates identify which display states (e.g., "color") should change in concert with other plots in the linkingGroup.

Usage

linking(
  linkingGroup = NULL,
  linkingKey = NULL,
  linkedStates = NULL,
  sync = NULL
)

Arguments

Value

a ggproto object

See Also

active, selection, zoom, hover, interactivity, l_getLinkedStates, l_setLinkedStates, l_configure

Examples

if(interactive() && requireNamespace("dplyr")) {
  h <- l_hist(mtcars$hp,
              linkingKey = rownames(mtcars),
              linkingGroup = "mtcars")

  mtcars %>%
    mutate(carName = rownames(mtcars)) %>%
    l_ggplot(mapping = aes(x = wt, y = hp, color = factor(cyl))) +
       geom_point(size = 4) +
       # push the states of scatter plot to the histogram
       linking(linkingGroup = "mtcars",
               linkingKey = ~carName,
               sync = "push")
}

loon.ggplot

Description

A bridge between loon widgets and gg objects. It can take either a loon widget, a gg object (ggplot, GGally::ggmatrix) or a l_ggplot object, then create a corresponding gg (or loon) graphics.

Usage

loon.ggplot(x, ...)

## S3 method for class 'gg'
loon.ggplot(x, ...)

## S3 method for class 'loon'
loon.ggplot(x, ...)

## S3 method for class 'zenplot'
loon.ggplot(x, ...)

## S3 method for class 'l_ggplot'
loon.ggplot(x, ...)

Arguments

Value

If the input is a ggplot object, the output would be a loon widget; conversely, if the input is a loon widget, then it returns a ggplot object. If it is a l_ggplot object, loon.ggplot helps to return a loon widget.

See Also

Richer examples are in loon2ggplot, ggplot2loon, l_ggplot

Examples

if(interactive()) {
######### loon --> gg #########
  # loon 3D plot
  l <- with(quakes,
    l_plot3D(long, lat, depth, linkingGroup = "quakes")
  )
  # equivalent to `loon2ggplot(l)`
  g <- loon.ggplot(l)
  g # a ggplot object

######### gg --> loon #########

  # ggplot histogram
  g <- ggplot(iris, mapping = aes(Sepal.Length, fill = Species)) +
    geom_histogram()
  # equivalent to `ggplot2loon(g)`
  l <- loon.ggplot(g)
  l # a loon widget

######### l_ggplot #########
  p <- l_ggplot(mpg, aes(displ, fill = factor(cyl))) +
       geom_histogram()
  class(p)
  # Function `print.l_ggplot` is called automatically
  p
  # Function `loon.ggplot` helps to return a loon widget
  q <- loon.ggplot(p)
  q
}

Deprecated functions in package loon.ggplot.

Description

The functions listed below are deprecated and will be defunct in the near future. When possible, alternative functions with similar functionality are also mentioned. Help pages for deprecated functions are available at help("geom_imageGlyph-deprecated"), help("geom_polygonGlyph-deprecated"), help("geom_pointrangeGlyph-deprecated"), help("geom_textGlyph-deprecated"), help("geom_serialAxesGlyph-deprecated"), help("lggplot-deprecated"), help("ggSerialAxes-deprecated")

Turn a loon widget to a ggplot object

Description

Create a ggplot object from a loon widget

Usage

## S3 method for class 'l_compound'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_facet_ggplot'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_facet_grid'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_facet_wrap'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_layer_graph'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_layer_histogram'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_layer_scatterplot'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_pairs'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_patchwork'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_serialaxes'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'zenLoon'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## Default S3 method:
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_plot'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_hist'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

## S3 method for class 'l_plot3D'
loon2ggplot(
  target,
  asAes = TRUE,
  selectedOnTop = TRUE,
  showNearestColor = FALSE,
  ...
)

Arguments

Details

In ggplot2, typically, there are two ways to set the aesthetic attributes, either take them as variables asAes = TRUE (set in the function aes()) or constants asAes = FALSE. The main benefits to consider them as variables are that 1. legend could be displayed; 2. convenient for further analysis.

In loon, when points were selected (highlighted), the order would be changed so that the highlighted points would be displayed at the front. To turn the loon plot static, if selectedOnTop = TRUE, the points would be partitioned into two groups – one group representing the un-highlighted points, and the other group representing the highlighted points. The un-highlighted group would be drawn first, then the selected group; if selectedOnTop = FALSE, no partition would be applied so that the displayed order remained. However, the highlighted points could be displayed at the back. See examples.

Value

a ggplot object (or a patchwork object, a extension of ggplot2)

Examples

if(interactive()) {
######## Basic ########
lp <- l_plot(iris,
             color = iris$Species,
             glyph = "circle")
gp <- loon2ggplot(lp)
gp # a ggplot object

# Add a smooth layer
# Reset initial plot object
gp$data <- gp$layers[[1]]$data
gp$mapping <- gp$layers[[1]]$mapping
gp +
  geom_smooth() +
  # give meaningful legend label names
  scale_color_manual(
    # make sure the order is correct
    values = unique(hex12tohex6(lp['color'])),
    labels = c("setosa", "versicolor", "virginica")
  )

# histogram
lh <- l_hist(mtcars$mpg,
             color = factor(mtcars$gear))

gh0 <- loon2ggplot(lh)
# facet by `fill`
gh0 + facet_wrap(~fill)

######## Argument `asAes` ########
gh1 <- loon2ggplot(lh, asAes = FALSE)
gh1
## Not run: 
# The bins are constructed by `ggplot2::geom_rect()`
# Very limited manipulations can be made
# ERROR will be returned
gh1 + facet_wrap(~fill)

## End(Not run)

######## Argument `selectedOnTop` ########
p <- l_plot(iris, color = iris$Species)
p['selected'][iris$Petal.Length > 5] <- TRUE
g <- loon.ggplot(p)
g
# facet by "Species"
g + facet_wrap(iris$Species)

# Something is wrong here. There is a pink point (at least one)
# in species "versicolor"! It is because after points are
# highlighted, the displayed order has been changed.
# Set `selectedOnTop` as FALSE, as in
loon.ggplot(p, selectedOnTop = FALSE) +
  facet_wrap(iris$Species)


######## l_patchwork --> ggplot ########
library(patchwork)
p1 <- ggplot(mtcars) + geom_point(aes(mpg, disp))
p2 <- ggplot(mtcars) +
   geom_boxplot(aes(gear, disp, group = gear))
p3 <- ggplot(mtcars) + geom_smooth(aes(disp, qsec))
design <- c(
  area(1,1),
  area(1,2),
  area(2,1,2,2)
)
pp <- p1 + p2 + p3 + plot_layout(design = design)

# turn a patchwork obj to a loon (l_compound)
lp <- ggplot2loon(pp)
# turn a loon (l_compound) back to a patchwork
plp <- loon2ggplot(lp)
plp # almost identical to pp


######## zneplots --> ggplot ########
library(zenplots)
stopifnot(packageVersion("zenplots") > "1.0.4")
zen <- zenplots::zenplot(iris, plot1d = "density", pkg = "loon")
ggzen <- loon.ggplot(zen)
ggzen +
  patchwork::plot_annotation(title = "This is a Patchwork!")

}

Transform geom layers to loon layers

Description

Function loonLayer is used to create loon non-interactive layers. For some ggplot2 extension packages, one can edit this function to realize the transformation.

Usage

loonLayer(
  widget,
  layerGeom,
  data,
  ggplotPanelParams,
  ggObj,
  parent,
  label,
  ...
)

Arguments

Explicitly draw plot

Description

Explicitly draw plot

Usage

## S3 method for class 'l_ggplot'
print(x, message = TRUE, ...)

Arguments

Value

Invisibly returns a loon widget

Box scaling in 3D rotation

Description

the variable is scaled to have equal ranges and, when center = TRUE, to be centred by the average of the min and max.

Usage

scaleBox(center = TRUE)

Arguments

Value

A trans object

See Also

l_scale3D

Position scales for continuous data (x, y & z)

Description

Scaling the coordinates for 3D visualization

Usage

scale_multi(trans = scaleBox(center = TRUE), ...)

Arguments

Details

In 3D rotation, different scales of variables x, y and z may cause an issue that the points appear to be off the window even with a minor tweak. Additionally, if one variable is in a large scale, the shape of the 3D plot may be dominated. Setting scale_multi can ensure the scales in the same measurement, as we rotate the plot, most points will stay inside the current view.

Value

a list of the ggproto objects

Examples

if(interactive()) {

dsamp <- dplyr::sample_n(diamonds, 100)

## Not run: 
# press `R`, then rotate with a minor tweak,
# Issues:
#   1: Points are displayed outside the window
#   2: Always in a line shape
l_ggplot(dsamp, aes(x = carat, y = price,
                    z = depth, colour = color)) +
  geom_point()
## End(Not run)

# set scales
l_ggplot(dsamp, aes(x = carat, y  =price,
                    z = depth, colour = color)) +
  geom_point() +
  scale_multi()

# customized `trans`
logp1_base10_trans <- scales::trans_new(
  name = "logp",
  trans = function(x) log(x + 1, base = 10),
  inverse = function(x) 10**x - 1,
  breaks = scales::log_breaks())

l_ggplot(dsamp, aes(x = carat, y = price,
                    z = depth, colour = color)) +
  geom_point() +
  scale_multi(trans = logp1_base10_trans)
}

selecting (deprecated)

Description

use selection, rather than selecting.

Usage

selecting(selected = NULL, selectBy = NULL, selectionLogic = NULL)

Arguments

Modify the selected component

Description

Set which elements (i.e., observations) are "selected". These will be shown as highlighted in the plot using the current "highlight" colour (see l_userOptions).

Usage

selection(selected = NULL, selectBy = NULL, selectionLogic = NULL)

Arguments

Details

There are two ways to directly select elements on, for example, a scatterplot using the mouse: either by "sweeping" or by "brushing". "Sweeping" allows us to sweep out a contiguous rectangular area of the plot, while, by "brushing", a fixed rectangular area is brushes across the plot selecting all points within the rectangle.

The selection logic give users more flexibility to users to not only highlight the elements, but also to downlight, and even to invert selections (changing the highlighted to downlighted, and vice versa).

Value

a ggproto object

See Also

active, linking, zoom, hover, interactivity, l_userOptions

Examples

if(interactive()) {

  # highlight the four gear cars
  fourGear <- rep(FALSE, nrow(mtcars))
  fourGear[mtcars$gear == 4] <- TRUE

  l_ggplot(mtcars, mapping = aes(x = wt, y = hp, color = factor(cyl))) +
    geom_point(size = 4) +
    # push the states of scatter plot to the histogram
    selection(selected = fourGear)
}

Zoom Plot Region

Description

Change the visible plot region by scaling to different elements of the display.

Usage

zoom(layerId = NULL, scaleToFun = NULL)

Arguments

Details

Argument layerId is used for additional plot region settings. If the layerId is set as NULL (default), the region of the interactive graphics loon will be determined by the ggplot object (i.e. coord_cartesian, xlim, etc); else one can use scaleToFun to modify the region of the layer.

The scaleToFun is a function to scale the region. If it is NULL (default), based on different layers, different scale functions will be applied. For example, if the layer is the main graphic model, i.e. l_plot l_hist, then the default scaleToFun is l_scaleto_plot; else if the layer is a general l_layer widget, the default scaleToFun would be l_scaleto_layer (see get_activeGeomLayers).

If it is not NULL, users can select one that precisely tailor their own problems. The table shows the available scaleToFun functions

Users can also supply their own function, providing its arguments match those of the functions shown in the above table.

Value

a ggproto object

See Also

active, linking, selection, hover, interactivity

Examples

if(interactive()) {
p <- l_ggplot(mtcars,
              mapping = aes(x = hp, y = mpg)) +
       geom_point(mapping = aes(color = factor(gear))) +
       geom_smooth(data = mtcars[mtcars$gear == 4, ],
                   method = "lm")
# a scatter plot with a fitted line on 4 gear cars
p
# scale to the second layer (smooth line)
p + zoom(layerId = 2)
# highlight the 3 gear cars
# scale to the selected points
p +
  selection(mtcars$gear == 3) +
  zoom(layerId = 1,
       scaleToFun = loon::l_scaleto_selected)
}