Symmetrical Transformation

Description

A1 and A2 are symmetrical on the two sides of Ax+By+C=0. The input of the function is A1, and the result is A2. The function also works when the line is horizontal or vertical. Note: the two shapes are symmetrical only when ggplot2::coord_fixed() is used.

Usage

ABCxy(
  x,
  A,
  B,
  C,
  p1 = NULL,
  p2 = NULL,
  f = NULL,
  group = TRUE,
  todf = TRUE,
  checks = TRUE
)

Arguments

Value

if todf = TRUE, the output will be a data frame with coordinates of possibly several polygons, otherwise, it will be a list of data frames. Data frames have 2 columns named "x" and "y", and if group = TRUE, a third column named "g" is added indicating group numbers.

Examples

library(ggplot2)
dat1=data.frame(x=c(0, 2, 2, 0), y=c(0, 0, 1, 1))
dat2=ABCxy(dat1, -1, -1, 3)
ggplot()+
	coord_fixed()+
	geom_polygon(data=dat1, aes(x=x, y=y), fill="red")+
	geom_polygon(data=dat2, aes(x=x, y=y), fill="blue")+
	geom_abline(intercept=3, slope=-1)		
dat3=ABCxy(dat1, p1=c(0, 1), p2=c(-0.5, 0), todf=TRUE)
ggplot()+
	coord_fixed()+
	geom_polygon(data=dat1, aes(x=x, y=y), fill="red")+
	geom_polygon(data=dat3, aes(x=x, y=y), fill="blue")+
	geom_abline(intercept=1, slope=2)

Generating Groups of Coordinates for Any Polygon

Description

Given your function to create a multiple of points (for example, points to form a polygon), this function generates x and y coordinates for groups of points of the same type with different parameters. The output of this function can be shown by ellipsexy and rectxy in this package.

Usage

ANYxy(myfun = NULL, ..., MoreArgs = NULL, group = TRUE, todf = TRUE)

Arguments

Examples

library(ggplot2)
# First, you need a function to generate
# x and y coordinates for a single group
# of points.
x_square=function(start, end, A, B){
	x=seq(start, end, 0.1)
	data.frame(x=x, y=A*(x^2)+B)
}
# All the arguments of your function 
# (here, start, end, A, B) should only accept
# vectors of length 1. And, the result of 
# your function should be a data frame
# of x and y coordinates 
# (here, coordinates of curves).
dat=ANYxy(myfun=x_square, 
	start=-1, end=1, A=c(1, 2), MoreArgs=list(B=1), 
	group=TRUE, todf=TRUE)
ggplot(dat)+geom_line(aes(x, y, group=g, color=factor(g)))

Adding Slash_n inside Strings

Description

This function simply adds change-line signs inside strings, so that they can be put vertically as the texts of x-axis.

Usage

add_slash_n(x, delete_space = TRUE, vertical_line = TRUE)

Arguments

Examples

lab=add_slash_n(c("a b-c", "d - ef ", "n"))

Layer for Drawing a Single Irregular Polygon with Shading Colors

Description

ggplot2::annotation_raster can only draw shading rectangles. However, this function can draw polygons of any shape with shading colors. See the shape argument and the raster argument.

Usage

annotation_shading_polygon(
  shape = data.frame(c(-1, 1, 0), c(0, 0, 1.732)),
  xmin = NULL,
  xmax = NULL,
  ymin = NULL,
  ymax = NULL,
  raster = NULL,
  interpolate = TRUE,
  result_interpolate = TRUE,
  shape_trim = NULL,
  raster_trim = NULL,
  result_trim = NULL,
  result = c("layer", "magick"),
  width = 800,
  height = NULL,
  res = 72
)

Arguments

Details

height can be used in the following ways:

• (1) an integer which will be directly passed to image::graph.

• (2) a character-like integer, e.g., height = "0.5". Suppose width = 400, the height that will be used is 400*0.5 = 200. This effectively prevents the image from becoming too large.

• (3) height = "coord_fixed". the ratio between height and width will be (top-bottom)/(right-left). And top, bottom, right and left are extreme values of shape when the latter is of class data.frame/matrix/gg.

• (4) height = "image". the width and height will be the width and height of raster when raster is a magick object.

• (5) height = NULL, the default. Now height is computed automatically. A ratio is computed first, ratio = (top-bottom)/(right-left). if the ratio is larger than 5 or smaller than 0.2, then height will be width*5 or width*0.2; else, the height will be treated in the same way as in (3) above. If shape is of class gg and it has uses coord_flip(), the height will be automatically adjusted. All these works are needed to prevent the image from becoming too large.

Examples

# Example 1
poly=ellipsexy(-1, 0, a=1, b=1)
m=matrix(rainbow(7))
ggplot()+
	coord_fixed()+
	annotation_shading_polygon(
		poly, raster=m
	)+
	annotation_shading_polygon(
		poly, raster=m, 
		xmin=1, xmax=5, 
		ymin=-1, ymax=1, 
	)
#
# Example 2, only an image
tt=annotation_shading_polygon(
	poly, result="magick", 
	width=280, height=280
)
#
# Example 3, both shape and raster are 
# ggplot plots.
p1=ggplot()+geom_tile(aes(x=1: 5, y=1: 5))
p2=ggplot()+geom_polygon(aes(x=c(0, 1, 1, 0), 
		y=c(0, 0, 1, 1)), fill="red")+theme_void()
ggplot()+coord_fixed()+
	annotation_shading_polygon(
		shape=p1, 
		xmin=1, xmax=10, 
		ymin=1, ymax=5, 
		raster=p2
	)

Layer for Transparent Text

Description

Suppose there is a colored rectangle with some texts and you want the texts to be transparent so that the colors of the background can be seen. Now you can use this function. The function can be used as a ggplot layer or a generator of image. NOTE: when the function is used as a layer, it uses ggplot2::annotation_raster to do the drawing, so you must set limits for the x axis and the y axis. See examples.

Usage

annotation_transparent_text(
  label,
  xmin,
  xmax,
  ymin,
  ymax,
  bg = "black",
  alpha = 0.5,
  operator = "out",
  interpolate = TRUE,
  result_interpolate = TRUE,
  expand = c(0.05, 0.05),
  family = "SimHei",
  fontface = 1,
  reflow = FALSE,
  place = "center",
  label_trim = NULL,
  bg_trim = NULL,
  result = c("layer", "magick"),
  width = 800,
  height = NULL,
  res = 72,
  ...
)

Arguments

Details

height can be used in the following ways:

• (1) an integer which will be directly passed to magick::image_graph.

• (2) a character-like integer, e.g., height = "0.5". Suppose width = 400, the height that will be used is 400*0.5 = 200. This effectively prevents the image from becoming too large.

• (3) height = "coord_fixed". the ratio between height and width will be (ymax-ymin)/(xmax-xmin).

• (4) height = "image". the width and height will be the width and height of bg when the latter is a magick object.

• (5) height = NULL, the default. Now height is computed automatically. If bg is a magick object, the width and height of the image will be used. If bg is not a magick object, a ratio is computed first, ratio = (ymax-ymin)/(xmax-xmin). if the ratio is larger than 5 or smaller than 0.2, then height will be width*5 or width*0.2; else, the height will be treated in the same way as in (3) above. All these works are needed to prevent the image from becoming too large.

Examples

# Example 1
m=matrix(rainbow(7), nrow=1)
ggplot()+coord_fixed()+
	xlim(0, 7)+ylim(-2, 4)+theme_void()+
	annotation_raster(
		raster=m, 
		xmin=0, ymin=-3, 
		xmax=7, ymax=5, 
		interpolate=TRUE
	)+
	annotation_transparent_text(
		label="R\nDATA\nVISUALIZATION", 
		xmin=0, xmax=7, 
		ymin=-1, ymax=3, 
		family="sans", fontface=2, alpha=0.8, 
		place="left", expand=c(0.08, 0.02)
	)
# 
# Example 2, this time the result is only an image.
tt=annotation_transparent_text(
	label="abcdefg", 
	xmin=1, xmax=8, 
	ymin=1, ymax=4, 
	alpha=0.6, 
	result="magick"
)
#
# Example 3, the rectangle is a matrix.
m=colorRampPalette(c("yellow", "purple"))(10)
ggplot()+coord_fixed(expand=FALSE)+
	theme(panel.background=element_rect(fill="red"))+
	annotation_transparent_text(
		label="hehehaha", 
		xmin=1, xmax=8, 
		ymin=1, ymax=4, 
		bg=m, alpha=1
	)
#
# Example 4, height is too large.
# Now you should explicitly set 
# width and height, otherwise, the 
# characters will become too flat.
x=c(0, 5, 10)
y=c(0, 500, 1000)
ggplot()+ylim(0, 4000)+
	geom_point(aes(x, y))+
	annotation_transparent_text(label="ha ha\nhe he", 
		xmin=0, xmax=10, ymin=1000, ymax=4000, bg="black", 
		width=300, height=150
	) # do not set height=NULL here

Counting Each Column and Summarizing in a Matrix

Description

This function counts the frequencies of each element of each column of a data frame or matrix. The frequencies of missing values and the 0 frequencies of non-existent values are also included in the final result.

Usage

count_each_column(x, answer = NULL, checks = TRUE)

Arguments

Examples

# values that do not appear in 
# the data can also be counted.
# a factor will be transformed into
# a character variable automatically.
x1=c("a", "b", "a", "b", NA)
x2=factor(x1)
x3=c("1", "3", "2", "1", "a")
dat=data.frame(x1, x2, x3, stringsAsFactors=FALSE)
res=count_each_column(dat, answer=c("c", "d", NA, "a"))
# logical value is OK.
x1=c(TRUE, TRUE, TRUE)
x2=c(TRUE, NA, NA)
dat=data.frame(x1, x2)
res=count_each_column(dat)
res=count_each_column(dat, c(TRUE, FALSE))

Generating Coordinates of Multiple Ellipses or Circles

Description

If radius a is equal to radius b, then the shape will be a circle. Note: the shapes are correct only when ggplot2::coord_fixed() is used.

Usage

ellipsexy(
  x = 0,
  y = 0,
  a = 2,
  b = 1,
  start = 0,
  end = 6.283185,
  angle = 0,
  n = 40,
  xytype = "middle",
  fan = FALSE,
  group = TRUE,
  todf = TRUE,
  checks = TRUE
)

Arguments

Value

if todf = TRUE, the output will be a data frame with coordinates of possibly several polygons, otherwise, it will be a list of data frames. Data frames have 2 columns named "x" and "y", and if group = TRUE, a third column named "g" is added indicating group numbers.

Examples

library(ggplot2)
dat1=ellipsexy(x=1, y=1, 
a=seq(1, 4, length.out=8), angle=seq(0, pi, length.out=8), 
	xytype="middleleft", n=30, todf=TRUE)		
ggplot()+coord_fixed()+
	geom_polygon(show.legend=FALSE, 
		data=dat1, aes(x=x, y=y, group=g, fill=factor(g)), alpha=0.3)

Enlarge a Color Matrix

Description

This is a convenient wrapper of colorRampPalette to enlarge a color matrix or raster.

Usage

enlarge_raster(x, n = c(10, 10), row_first = TRUE, space = "rgb")

Arguments

Examples

library(ggplot2)
# the original matrix
m=matrix(c(
	"red", "yellow", "green", 
	"blue", "purple", "cyan", 
	"black", "orange", "grey"), byrow=TRUE, nrow=3)
# enlarge the matrix
mm=enlarge_raster(m, c(15, 15), space="Lab")
ggplot()+xlim(0, 10)+ylim(0, 5)+coord_fixed()+
	annotation_raster(mm, 
		xmin=0, xmax=10, ymin=0, ymax=5, interpolate=TRUE)

Geom Layer for Circle with Absolute Size

Description

This layer uses centimeter as unit to draw circles so that the size and shape will not be influenced by the change of the coordinate systems (even when a polar system is used). Note: this function does not have linetype and n arguments.

Usage

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

Arguments

Details

Accepted properties are:

• (1) rcm radius in centimeter.

• (2) color color of the outline.

• (3) fill color inside the shape.

• (4) alpha alpha of color and fill.

• (5) size line width of the outline.

• (6) x x coordinates of the middle points.

• (7) y y coordinates of the middle points.

Examples

library(ggplot2)
dat=data.frame(x=1: 10, y=rep(5, 10), R=rep(c(0.5, 1), 5))
ggplot(dat)+xlim(0, 11)+ylim(1, 9)+
	geom_circle_cm(aes(x=x, y=y, fill=factor(R)), rcm=dat$R, alpha=0.5)

Geom Layer for Ellipse with Absolute Size

Description

This layer uses centimeter as unit to draw ellipse so that its size and shape will not be influenced by the coordinate systems (even when a polar system is used).

Usage

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

Arguments

Details

Accepted properties are:

• (1) rcm radius in centimeter.

• (2) ab it means to what extent radius a of an ellipse is larger than radius b. However, its true meaning is the aspect ratio which is used by gridExtra::ellipseGrob and indicates the extent to which y dimension is flattened. So, say, when ab = 2, radius a is larger than b, but it is not exactly 2 times larger.

• (3) color color of the the outline.

• (4) fill color inside the shape.

• (5) alpha alpha of color and fill.

• (6) size line width of the outline.

• (7) linetype line type.

• (8) angle angle of rotation from 0 degree and in anti-clockwise direction.

• (9) n the number of points to draw the shape. Note: it must be written inside the aes(...) function.

• (10) x x coordinates of middle points.

• (11) y y coordinates of middle points.

Examples

library(ggplot2)
dat=data.frame(x=c(1, 3, 5, 7, 9), y=rep(5, 5)) 
ggplot(dat)+xlim(0, 11)+ylim(1, 9)+
	geom_ellipse_cm(aes(x=x, y=y), fill="red", ab=seq(1, 4, length.out=5))  
ggplot(dat)+xlim(0, 11)+ylim(1, 9)+
	geom_ellipse_cm(aes(x=x, y=y, fill=factor(x)), ab=3, angle=c(0, pi/4, pi/3, pi/2, 0.75*pi))

Geom Layer for Drawing Multiple Rasters

Description

Unlike annotation_raster which draws only 1 raster, this layer draws one or more rasters at the same time. The data must be a tbl object created by package tibble and the reason is that, as we must give each rectangle a vector of colors, the column that contains these vectors of colors must be a list rather than a vector. A list can be a column for tbl object, not for a normal data frame. See examples. Accepted properties are:

• (1) xmin.

• (2) xmax.

• (3) ymin.

• (4) ymax.

• (5) raster. a list with 1 or more rasters. If you have only 1 raster, you also have to put it into a list. Each raster should be a matrix, a raster object, a character vector or a magick object read into R by magick::image_read. You can also use a data frame created by package tibble to combine xmin, xmax, ymin, ymax, raster.

• (6) interpolate. It is the same as that in annotation_raster except that the default value is TRUE. It can be used either inside or outside the aes(...) function. Its length must be either 1 or the same as the number of rasters.

• (7) flip. The default is FALSE. You only need to use TRUE when you use coord_flip. Used outside the aes(...) function.

Usage

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

Arguments

Examples

# Example 1: use vectors and a list. 
mycolor=list(
	c1=matrix(c("red", "blue", "green", "yellow"), nrow=2), 
	c2=matrix(c("green", "yellow")), 
	c3=matrix(c("purple", "red")))
xmin=1: 3
xmax=(1: 3)+0.8 
ymin=c(0, 1, 2)
ymax=c(1, 3, 5)
ggplot()+
	geom_multi_raster(aes(xmin=xmin, xmax=xmax, 
	ymin=ymin, ymax=ymax, raster=mycolor))
#
# Example 2: the same as example 1
# except flip=TRUE.
ggplot()+coord_flip()+
	geom_multi_raster(aes(xmin=xmin, xmax=xmax, 
	ymin=ymin, ymax=ymax, raster=mycolor), flip=TRUE)

Geom Layer for Rectangle with Absolute Size

Description

This layer uses centimeter as unit to draw rectangles so that the size and shape will not be influenced by the coordinate systems (even when a polar system is used).

Usage

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

Arguments

Details

Accepted properties are:

• (1) width width in centimeter.

• (2) height height in centimeter.

• (3) color color of the outline.

• (4) fill color inside the shape.

• (5) alpha alpha of color and fill.

• (6) size line width of outline.

• (7) linetype line type.

• (8) hjust horizontal adjustment, default is 0.5 which means no adjustment.

• (9) vjust vertical adjustment, default is 0.5 which means no adjustment.

• (10) x x coordinates of middle points.

• (11) y y coordinates of middle points.

Examples

library(ggplot2)
ggplot()+xlim(-0.5, 10.5)+
	geom_rect_cm(aes(x=1: 10, y=rep(4, 10)), fill="red", height=rep(1: 2, each=5), 
		vjust=rep(c(0, 0.5), 5))+
	geom_point(aes(x=1: 10, y=rep(4, 10)), color="green")

Geom Layer for Drawing Shading Barplot

Description

This function is similar to geom_bar(aes(x, y), stat="identity") except that it draws bars with shading colors. Unlike gg_shading_bar which is a convenient function, this function is used as a ggplot layer. Accepted properties are different from those in geom_multi_raster and gg_shading_bar.

• (1) x. It is the same as that in geom_bar.

• (2) y. It is the same as that in geom_bar.

• (3) raster. It should be a list with 1 or more character vectors of colors. If the list only has 1 vector, all the bars will use the same shading pattern. If you have, for example, 5 bars to draw, then you have to put 5 vectors of colors into a list. If you use a data frame, it must be a data frame made by package tibble, and the column for raster should be a list.

• (4) width. It is the same as that in geom_bar.

• (5) flip. The default is FALSE. You only need to use TRUE when you use coord_flip. Use outside the aes(...) function.

• (6) modify_raster. If it is TRUE (default), colors will be smoothed using the value of smooth. If raster has enough colors, you can set this to FALSE. It is the same as that in gg_shading_bar.

• (7) equal_scale. The default is FALSE. When it is TRUE, a bar will use a certain part of the shading colors according to a global scale. It is the same as that in gg_shading_bar.

• (8) smooth. The default is 15. The number of shading colors each bar has. The bigger, the better. It is the same as that in gg_shading_bar.

• (9) space. The color space that is used. It can be "rgb" (default) or "Lab".

• (10) orientation. This parameter mimics the same parameter used in geom_bar, though acts differently. This enables to flip the x axis and y axis without using coord_flip. If it is NA or "x" (default), it supposes x = SOME LABELS and y = SOME VALUES. If it is "y", you must set x = SOME VALUES and y = SOME LABELS. These effects are the same as geom_bar.

NOTE: the function does interpolation as default, so you does not need to use interpolate parameter. And, unlike gg_shading_bar, this function does not draw lines around rectangles.

Usage

geom_shading_bar(
  mapping = NULL,
  data = NULL,
  stat = "identity",
  position = "identity",
  na.rm = FALSE,
  show.legend = NA,
  inherit.aes = TRUE,
  width = 0.9,
  flip = FALSE,
  modify_raster = TRUE,
  smooth = 15,
  equal_scale = FALSE,
  space = "rgb",
  orientation = "x",
  ...
)

Arguments

Examples

# Example 1: use vectors.
x=c("b", "a", "c", "d", "e")
y=c(2, 1, 3, 5, 4)
raster=list(c("blue", "red"), c("green", "orange"), 
	c("cyan", "yellow"), c("purple", "orangered"), c("grey", "red"))
ggplot()+
	geom_shading_bar(aes(x=x, y=y, raster=raster), smooth=40)
#
# Example 2: other parameters
x=1: 5
y=c(1, 2, -3, 5, 4)
raster=list(c("blue", "red"))
ggplot()+
	geom_shading_bar(aes(x=x, y=y, raster=raster), 
		smooth=50, width=0.6, equal_scale=TRUE)+
	scale_x_continuous(breaks=1: 5, labels=letters[1: 5])

Obtaining the Colors of Positions Clicked

Description

The function draws an image and asks the user to click on the positions whose colors the user wants to know. NOTE: after clicking, you must press Esc button to continue. The result is a vector of colors in hex mode.

Usage

get_click_color(x)

Arguments

Checking Min, Max, Labels and Label Positions

Description

Given a numeric vector or a ggplot object, the function will check the range, labels and label positions (the same as major grid lines) that will used on the axis. The result is a length 5 list for min limit, max limit, labels, major grid-line positions, all (major and minor) grid-line positions.

Usage

get_gg_label(
  a = NULL,
  b = NULL,
  v = NULL,
  gg = NULL,
  mult = 0.05,
  add = 0,
  axis = "y"
)

Arguments

Examples

get_gg_label(a=1, b=1000)
# The following three have the same results.
get_gg_label(a=1, b=1000, mult=0)
get_gg_label(v=c(1, 500, 1000), mult=0)
p=ggplot()+geom_point(aes(1: 3, c(1, 500, 1000)))+
  scale_y_continuous(expand=expansion(mult=0))
get_gg_label(gg=p)

Drawing Barplot with Shading Colors

Description

In ordinary barplot, each bar has only one color. This function aims to draw a barplot whose bars have shading effect. Note: unlike ggplot2::geom_bar, this function can only deals with a vector of frequencies.

Usage

gg_shading_bar(
  v,
  labels = NULL,
  raster = NULL,
  flip = FALSE,
  change_order = "normal",
  equal_scale = FALSE,
  smooth = 15,
  interpolate = TRUE,
  width = 0.8,
  color = NA,
  linetype = 1,
  size = 1,
  modify_raster = TRUE,
  space = "rgb",
  ...
)

Arguments

Examples

library(ggplot2)
x=c(10, 30, 25, 6)
lab=c("children", "youth", "middle", "aged")
r=list(c("cyan", "red"), c("blue", "yellow"), 
	c("green", "orange"), c("grey", "black"))
#
## (1) change_order
# change_order = "ordinary", the default
p1=gg_shading_bar(v=x, labels=lab)
# change_order = "big"
p2=gg_shading_bar(v=x, labels=lab, change_order="big")
# flip and let the largest on the top
p3=gg_shading_bar(v=x, labels=lab, 
	change_order="small", flip=TRUE)
#
## (2) how to use argument raster
p1=gg_shading_bar(v=x, labels=lab, raster=r)
p2=gg_shading_bar(v=x, labels=lab, raster=c("green","red"))
#
## (3) how to use argument equal_scale
# equal_scale = FALSE
# the far side of each bar is red
gg_shading_bar(c(3, 5), raster=c("green", "red"))
# equal_scale = TRUE
# the far side of the shorter bar 
# is not red. Rather, it is something
#' between red and green
gg_shading_bar(c(3, 5), raster=c("green", "red"), 
	equal_scale=TRUE)

Colorize an Image according to Gray Scale

Description

A color image can be converted to one with different degrees of gray. Then, colors in a palette can be added according to the gray degrees. The function is a simple wrapper of scales::col_numeric. The pixels which are deliberately assigned "transparent" in the original magick image will always kept unchanged.

Usage

image_col_numeric(
  x,
  palette = c("purple", "yellow"),
  n = 256,
  alpha = FALSE,
  result = "magick",
  res = 144
)

Arguments

Cut out a Subregion of an Image by Mouse Click

Description

This function is a wrapper of magick::image_crop. While the latter asks you to set a geometry parameter, this function enables you to set the four sides of a subregion only by click the mouse. You must click at least 2 times (that is, click on 2 different points to define a rectangle). After clicking, please press Esc on your keyboard. You can also designate an irregular polygon by mouse with at least 3 clicks. If it is irregular, you MUST click on positions in order (something like that, when you draw a polygon in R, you must input the positions of points in order).

Usage

image_crop_click(x, only_value = FALSE, rectangle = TRUE, trim = FALSE)

Arguments

Keep Some Colors Unchanged and Make Others into Grayscale

Description

This function keeps pixels with certain colors unchanged and transforms others into grayscale. The function is in fact a wrapper of magick::image_transparent, so it uses the latter's color and fuzz parameters. NOTE: the function only works for fully opaque or fully transparent (labelled as "transparent") pixels.

Usage

image_keep_color(x, color = NULL, fuzz = 10, result = "magick")

Arguments

Get the Width and Height of the Mouse Clicked Points

Description

This function simply gets the width and height values of the points on which you click. The result is a list of two vectors, the first vector is for width, the second for height.

Usage

image_locator(x, rectangle = FALSE)

Arguments

Modify the H, S, V Values of a Color Vector or an Image

Description

The function modifies the H (0 - 1), S, V values of a vector of colors or an image. The three channels can be modified separately. However, the most frequently used is only the V modification. The ways to modify include: setting values to some specified values (set_*), adding (add_*), multiplying the original values (mult_*), rescaling the original values (rescale_*), using a function to recompute values (fun_*). The most useful way is to use some internal curves that mimic those PS-like apps. DO see Details.

Usage

image_modify_hsv(
  x,
  set_h = NULL,
  add_h = NULL,
  mult_h = NULL,
  rescale_h = NULL,
  fun_h = NULL,
  set_s = NULL,
  add_s = NULL,
  mult_s = NULL,
  rescale_s = NULL,
  fun_s = NULL,
  set_v = NULL,
  add_v = NULL,
  mult_v = NULL,
  rescale_v = NULL,
  fun_v = NULL,
  result = "magick",
  res = 144
)

Arguments

Details

fun_* can be a function or a named list which tells the function which internal function is to be used. You must ensure values used by the function specified by you to be in the range [0, 1] for H, S, V modification and [0, 255] for R, G, B modification. Also, you'd better make sure the output values of the function are in

When fun_* is a list, it should be written in the following way:

• (1) fun_* = list(fun = "s", c1 = -2, c2 = 2, domain = c(0, 1)) An "s" curve will be used. c1 points out how to deal with values below 0.5, c2 with values above 0.5. For c1 and c2, a value larger than 0 means a curvature towards y = 1, and a value smaller than 0 means a curvature towards y = 0. So, c1 < 0 and c2 > 0 will make an s shape curve. c1 and c2 can be any number, though those with absolute values below 4 are quite good (default is -2 and 2). 0 means no change. domain specifies the value domain to put the result. The default is c(0, 1) which means not to rescale, thus 0.1 is 0.1. However, if you set domain = c(0.5, 1), then 0.1 will be 0.55. If you do not know how to set domain, just ignore it.

• (2) fun_* = list(fun = "circle", value = 0.5) When the fun is "circle" or "c", an arc will be used. value must be a number between -1 and 1 (default is 0.5). A number larger than 0 means the curvature is towards y = 1, and a number smaller than 0 means it is towards y = 0. value should not be 0.

• (3) list(fun_* = "linear", x0 = 0.4, y0 = 0.6) This makes a linear modification except that there is a breakpoint. The default point is (0.4, 0.6) which means: suppose all the original numbers and output numbers are in the [0, 1] range and the points with their x position smaller than 0.4 will be put along the line that links (0, 0) and (0.4, 0.6), and, those with x position larger than 0.4 will be put along the line that links (0.4, 0.6) and (1, 1).

Examples

# First create an image
library(magick)
mycolor=grDevices::hsv(0, s=seq(0.1, 0.9, 0.1), 
	v=seq(0.1, 0.9, 0.1))
img=image_graph(width=400, height=400)
print(showcolor(mycolor)+theme_void())
dev.off()
# Now increase S values with 
# an internal circle curve and 
# set V values between [0.5, 1].
res=image_modify_hsv(img, 
	fun_s=list("circle", value=1), 
	rescale_v=c(0.5, 1))

Modify Only a Subregion of an Image

Description

The function allows you to modify a subregion of your image (or, the opposite, keep the subregion unchanged while modifying other parts). You can set the four sides of the subregion or an irregular polygon by mouse click. If it is irregular, you MUST click in order.

Usage

image_modify_local(
  x,
  FUN,
  geometry = "click",
  local = "local",
  rectangle = TRUE,
  trim = FALSE,
  ...
)

Arguments

Modify both a Subregion and the Whole of an Image

Description

The function is similar to image_modify_local but with different parameters. It modifies both a subregion of the image and the whole image, and then combines them. The subregion can be chosen either by numeric values or by mouse click ,which is the same as image_modify_local.

Usage

image_modify_local2(x, FUN1, FUN2 = NULL, geometry = "click", rectangle = TRUE)

Arguments

Modify R, G, B Values of an Image

Description

The function modifies the R, G, B values of an image and is used in the same way as image_modify_hsv in this package. The three channels can be modified separately. The ways to modify include: setting values to some specified values (set_*), adding (add_*), multiplying the original values (mult_*), rescaling the original values (rescale_*), using a function to recompute values (fun_*). The most useful way is to use some internal curves that mimic those PS-like apps.

Usage

image_modify_rgb(
  x,
  set_r = NULL,
  add_r = NULL,
  mult_r = NULL,
  rescale_r = NULL,
  fun_r = NULL,
  set_g = NULL,
  add_g = NULL,
  mult_g = NULL,
  rescale_g = NULL,
  fun_g = NULL,
  set_b = NULL,
  add_b = NULL,
  mult_b = NULL,
  rescale_b = NULL,
  fun_b = NULL,
  result = "magick",
  res = 144
)

Arguments

Details

Several internal curves can be used. Please see the Details part of image_modify_hsv.

Modify R, G, B Values according to V values

Description

While the image_modify_rgb function modifies R, G, B with reference to the original values, image_modify_rgb_v also takes into account the brightness (V) values. It is similar to those apps which divide an image into a bright part and a dark part (and, for example, you can increase red in the bright part and decrease red in the dark part.

Usage

image_modify_rgb_v(
  x,
  fun_r = NULL,
  fun_g = NULL,
  fun_b = NULL,
  alpha = FALSE,
  rescale_v = NULL,
  result = "magick",
  res = 144
)

Arguments

Details

This function uses custom functions or internal curves to make modification. See the Details part of image_modify_hsv to know how to use them. Note: values will be coerced to be in the [0, 255] range with no warning. For example, the original value is 240 and it becomes 280 in the output, then it will be set to 255 automatically.

Keep Certain Colors Unchanged and Make Others Transparent

Description

This function is an inverse version of magick::image_transparent. While the latter makes certain colors transparent, the former keeps them unchanged and make others transparent.

Usage

image_transparent_inverse(x, color, fuzz = 0)

Arguments

Combine a Matrix of Colors and a Matrix of Alpha Values

Description

The function is a wrapper of scales::alpha. While the latter only works on vectors, the former can combine a matrix of colors and a matrix of alpha values as long as the two have the same numbers of rows and columns.

Usage

raster_alpha(color, alpha, result = "raster", res = 144)

Arguments

Examples

# A color matrix
co=c("red", "yellow", "green", "blue")
co=rbind(co, co, co)
# An alpha matrix
alp=c(1, 0.6, 0.3, 0.1)
alp=rbind(alp, alp, alp)
# Now combine the two
result=raster_alpha(co, alp)

Generating Coordinates of Multiple Rectangles

Description

Note: the shapes are correct only when ggplot2::coord_fixed() is used.

Usage

rectxy(
  x = 0,
  y = 0,
  a = 1,
  b = 1,
  angle = 0,
  xytype = "middle",
  group = TRUE,
  todf = TRUE,
  checks = TRUE
)

Arguments

Value

if todf = TRUE, the output will be a data frame with coordinates of possibly several polygons, otherwise, it will be a list of data frames. Data frames have 2 columns named "x" and "y", and if group = TRUE, a third column named "g" is added indicating group numbers.

Examples

library(ggplot2)
dat1=rectxy(x=4, y=3, a=2, b=1, angle=0, xytype="bottomleft", todf=TRUE) 
dat2=rectxy(x=4, y=3, a=2, b=1, angle=pi/6, xytype="bottomleft", todf=TRUE)
ggplot()+
	geom_polygon(data=dat1, aes(x=x, y=y), fill="red", alpha=0.3)+
	geom_polygon(data=dat2, aes(x=x, y=y), fill="blue", alpha=0.3)+	
	coord_fixed()

Resize an Image According to the Other Image or to Ratios

Description

Simple wrapper of magick::image_resize. See the parameters below.

Usage

resize_to_standard(x, standard = 0.5, what = "all", scale = TRUE)

Arguments

Rotation Transformation

Description

A2 (output) is the result of rotating A1 (input) around a point. Note: the two shapes look the same (though with different angles) only when ggplot2::coord_fixed() is used.

Usage

rotatexy(
  x,
  angle = pi/4,
  xmiddle = 0,
  ymiddle = 0,
  f = NULL,
  group = TRUE,
  todf = TRUE,
  checks = TRUE
)

Arguments

Value

if todf = TRUE, the output will be a data frame with coordinates of possibly several polygons, otherwise, it will be a list of data frames. Data frames have 2 columns named "x" and "y", and if group = TRUE, a third column named "g" is added indicating group numbers.

Examples

library(ggplot2)
dat1=data.frame(x=c(0, 4, 4, 0), y=c(0, 0, 2, 2))
dat2=data.frame(x=c(5, 6, 6, 5), y=c(4, 4, 8, 8))
dat3=rotatexy(list(dat1, dat2), angle=c(pi, pi/4), 
	xmiddle=c(0, 5), ymiddle=c(0, 4), todf=TRUE)
ggplot()+
	coord_fixed()+
	geom_polygon(data=dat1, aes(x=x, y=y), fill="red", alpha=0.2)+
	geom_polygon(data=dat2, aes(x=x, y=y), fill="blue", alpha=0.2)+
	geom_polygon(show.legend=FALSE, data=dat3, 
		aes(x=x, y=y, group=g, fill=factor(g)), alpha=0.2)

Converting Numeric Values into Characters with the Same Digits

Description

This simple function is to facilitate something like decimal horizontal adjustment which demands each value has the same digits after the decimal point.

Usage

round_text(x, digits = 2, na = NULL)

Arguments

Examples

v=c(3, 3.1, 3.456, 3.452, 3.77, NA, 0, 10.56332)
res=round_text(v, 2, na=NA)

Scale values into a Certain Location

Description

A simple function to put numeric values into a certain interval. Suppose you have 20, 60, 80, 100, and you want them to be in the interval of [0, 1], so you can get 0, 0.5, 0.75, 1.

Usage

scale_free(
  x,
  left = 0,
  right = 1,
  reverse = FALSE,
  xmin = NULL,
  xmax = NULL,
  na.rm = FALSE
)

Arguments

Examples

y=scale_free(c(-1, 0, 2))
y=scale_free(c(-1, 0, 2), rev=TRUE)
#
# x is a data frame.
x=data.frame(
	c(-1, 0, 0, 0, 2), c(-1, 0, 0, 0, 2), 
	c(-2, 0, 2, 4, 6), c(-2, 0, 2, 4, 6)
)
y=scale_free(x, 
	left=0, right=10, 
	reverse=c(FALSE, TRUE, FALSE, TRUE)
)
y=scale_free(x, 
	left=c(0, 0, 100, 100), right=c(10, 100, 200, 200), 
	reverse=c(FALSE, TRUE, FALSE, TRUE)
)

Create a Shading Raster with a Palette

Description

The function is a simple wrapper of scales::col_numeric. The function creates a matrix of colors that can be used to draw a shading rectangle. There are 2 ways to use the function, see the following parameters.

Usage

shading_raster(
  nr = NULL,
  nc = NULL,
  middle = NULL,
  palette = c("blue", "red"),
  mat = NULL,
  FUN = NULL
)

Arguments

Examples

# Use method 1.
r=shading_raster(nr=31, nc=60, middle=c(10, 55), 
	palette=c("darkorange", "red", "purple"))
ggplot()+xlim(0, 8)+ylim(0, 6)+
	annotation_raster(r, xmin=-Inf, xmax=Inf, 
	ymin=-Inf, ymax=Inf, interpolate=TRUE)
 # Use method 2.
r=matrix(c(
	1, 2, 3, 4, 5, 6, 7, 8, 
	1, 2, 3, 4, 5, 6, 7, 8, 
	1, 1, 1, 1, 1, 1, 1, 1), 
	nrow=3, byrow=TRUE)
r=shading_raster(mat=r, palette=c("green", "blue"))

Show a Color Palette

Description

Simple function to show colors. NOTE: do not add coord_flip().

Usage

showcolor(x, label_size = 15, ...)

Arguments

Examples

# A palette used by David Hockney
co=c("#833822", "#C03800", "#D3454C", 
	"#DC6A30", "#F29856", "#FEEF70", 
	"#A5D56D", "#16D670", "#00932F", 
	"#03592E", "#04B7B0", "#007BA9", 
	"#EC46BF", "#6A2C8F"
)
showcolor(co, label_size=10)

Reordering Points to Form a "s" Shape

Description

This is a convenient function to generate points with x and y coordinates (which form a 2-column data.frame). It is much like expand.grid. The points generated by expand.grid always in this "s" order: the bottom line, form left to right, and the second line, from left to right. However, spathxy allows you choose the order you want. See examples.

Usage

spathxy(
  x,
  y,
  first = "right",
  second = "top",
  change_line = FALSE,
  stringsAsFactors = TRUE
)

Arguments

Value

always a 3-column data frame. Column x and y are coordinates of points; column index contains the index number of points.

Examples

library(ggplot2)
#
# dat1 is generated by expand.grid
# Note the difference between dat1 and dat2.
# dat3 is the same as dat1.
dat1=expand.grid(1: 3, 1: 7)
colnames(dat1)=c("x", "y")
dat2=spathxy(1: 3, 1: 7, 
	change_line=FALSE, first="right", second="top")
dat3=spathxy(1: 3, 1: 7, 
	change_line=TRUE, first="right", second="top")
#
mycolor=rainbow(nrow(dat1), end=0.6)
ggplot(dat1)+geom_path(aes(x, y), color=mycolor, size=3)
ggplot(dat2)+geom_path(aes(x, y), color=mycolor, size=3)
ggplot(dat3)+geom_path(aes(x, y), color=mycolor, size=3)

Stretching Transformation

Description

A2 (output) is the result of enlarging (or shrinking) A1 (input) in x dimension and y dimension. Note: the two shapes manifest enlarging or shrinking effect only when ggplot2::coord_fixed() is used.

Usage

stretchxy(
  x,
  xlarge = 2,
  ylarge = 2,
  f = NULL,
  group = TRUE,
  todf = TRUE,
  checks = TRUE
)

Arguments

Value

if todf = TRUE, the output will be a data frame with coordinates of possibly several polygons, otherwise, it will be a list of data frames. Data frames have 2 columns named "x" and "y", and if group = TRUE, a third column named "g" is added indicating group numbers.

Examples

library(ggplot2)
dat1=data.frame(x=c(0, 1, 1), y=c(0, 0, 1))
dat2=data.frame(x=c(4, 5, 5, 4), y=c(0, 0, 3, 3))
dat3=stretchxy(list(dat1, dat2), xlarge=3, ylarge=c(3, 2), todf=TRUE)
ggplot()+coord_fixed()+
	geom_polygon(data=dat1, aes(x, y), fill="red", alpha=0.3)+
	geom_polygon(data=dat2, aes(x, y), fill="blue", alpha=0.3)+
	geom_polygon(data=dat3, aes(x, y, fill=g, group=g), fill="blue", alpha=0.3)	

Generating Lines Which Link One Points to Many

Description

Suppose there is a middle point a, this function simultaneous generates points on lines that start from a to other points.

Usage

sunshinexy(
  x = 0,
  y = 0,
  outer = data.frame(1, 1),
  n = 10,
  delete_n = 0,
  distance = FALSE,
  checks = TRUE
)

Arguments

Value

A data frame that has 3 columns. The first and second columns are named "x" and "y", the third column is named "g" indicating group numbers. If distance = TRUE, a fourth column is added which indicates the distances from the middle point to other points.

Examples

library(ggplot2)
p=c(1, 1, 0, -1, -1, -1, 0, 1)
q=c(0, 1, 1, 1, 0, -1, -1, -1)
pq=data.frame(cbind(p, q))
dat=sunshinexy(outer=pq, n=20, delete_n=5, distance=TRUE)
ggplot()+coord_fixed()+theme_void()+
	geom_point(data=pq, aes(p, q), size=4)+
	geom_line(show.legend=FALSE, data=dat, aes(x, y, group=g, color=distance), size=2)+
	scale_color_continuous(low="blue", high="red")

Simple Text ".gif" File

Description

This is a wrapper of functions in package ggfittext and magick. The output is a ".gif" with changing texts and colors. Characters are automatically enlarged or shrunk.

Usage

textgif(
  text,
  text_color = NULL,
  bg_color = NULL,
  reflow = FALSE,
  width = 200,
  height = 100,
  family = "SimHei",
  fontface = 1,
  fps = 2,
  output = NULL,
  ...
)

Arguments

Examples

mytext=c("AAA", "BBB", "CCC")
color1=c("orange", "red", "white")
color2=c("black", "blue", "green")
g1=textgif(mytext, text_color=color1, bg_color=color2, 
	width=180, height=120, fps=2, family="serif")