将自定义图像显示为geom_point

Question

是否可以在R ggplot中将自定义图像(例如png格式)显示为geom_point？

library(png)
pic1 <- readPNG("pic1.png")

png("Heatmap.png", units="px", width=3200, height=3200, res=300)
ggplot(data_frame, aes(medium, day, fill = Transactions))  +
   geom_tile(colour="white")  +
   facet_grid(dime3_year~dime3_month) + 
   scale_fill_gradient(high="blue",low="white") +
   theme_bw() + 
   geom_point(aes(dime3_channel, day, size=Conv,alpha=Conv,image=(annotation_raster(pic1,xmin=0,ymin=0,xmax=5,ymax=5)),color="firebrick")) +

给出错误:

不知道如何自动选择proto/environment类型的对象的比例.默认为连续错误:美学必须是长度为1或与dataProblems相同的长度:( annotation_raster(conv_pic,xmin = 0,ymin = 0,xmax = 5,ymax = 5))

Answer 1

这并不是你想要的,geom_point但它可能暗示了一个快速的选择.但是,它包含了相当粗略的调整x和y限制.

library(png)
library(ggplot2)

img <- readPNG(system.file("img", "Rlogo.png", package="png"))

ggplot(mtcars, aes(mpg, wt)) + 
       geom_blank() +
       mapply(function(xx, yy) 
          annotation_raster(img, xmin=xx-1, xmax=xx+1, ymin=yy-0.2, ymax=yy+0.2),
          mtcars$mpg, mtcars$wt) 

对于方面,请参阅Kohske关于如何改变mapply功能的答案.

编辑

我认为这实际上可以更好地使用annotation_custom(),就像Deb的回答一样.下面允许遍历所有点,而不是必须使用单独的annotation_custom调用.从上面稍微改变一下,grob似乎需要重命名(来自链接的评论)

g <- rasterGrob(img, interpolate=FALSE)

ggplot(mtcars, aes(mpg, wt)) + 
          geom_blank() +
          mapply(function(xx, yy, ii) {
          g$name <- ii
          annotation_custom(g, xmin=xx-1, xmax=xx+1, ymin=yy-0.2, ymax=yy+0.2)},
          mtcars$mpg, mtcars$wt, seq_len(nrow(mtcars))) 

Answer 2

点geom用于创建散点图,并且似乎不是设计用于执行所需的操作,即显示自定义图像.但是,这里回答了类似的问题,表明问题可以通过以下步骤解决:

(1)读取要显示的自定义图像,

(2)使用rasterGrob()函数渲染给定位置,大小和方向的栅格对象,

(3)使用绘图功能,如qplot(),

(4)使用geom,例如用作annotation_custom()静态注释,指定user206提到的x和y限制的粗略调整.

使用下面的代码,我可以得到两个自定义图像img1.png和img2.png位于给定的xmin,xmax,ymin和ymax.

library(png)
library(ggplot2)
library(gridGraphics)
setwd("c:/MyFolder/")

img1 <- readPNG("img1.png")
img2 <- readPNG("img2.png")
g1 <- rasterGrob(img1, interpolate=FALSE)
g2 <- rasterGrob(img2, interpolate=FALSE)
qplot(1:10, 1:10, geom="blank") + 
  annotation_custom(g1, xmin=1, xmax=3, ymin=1, ymax=3) +
  annotation_custom(g2, xmin=7, xmax=9, ymin=7, ymax=9) +  
  geom_point()

Answer 3

DL Miller提供了另一种解决方案ggproto().https://github.com/dill/emoGG

library(ggplot2)
library(grid)
library(EBImage)
img <- readImage(system.file("img", "Rlogo.png", package = "png"))
RlogoGrob <- function(x, y, size, img) {
    rasterGrob(x = x, y = y, image = img, default.units = "native", height = size, 
        width = size)
}

GeomRlogo <- ggproto("GeomRlogo", Geom, draw_panel = function(data, panel_scales, 
    coord, img, na.rm = FALSE) {
    coords <- coord$transform(data, panel_scales)
    ggplot2:::ggname("geom_Rlogo", RlogoGrob(coords$x, coords$y, coords$size, 
        img))
}, non_missing_aes = c("Rlogo", "size"), required_aes = c("x", "y"), default_aes = aes(size = 0.05), 
    icon = function(.) {
    }, desc_params = list(), seealso = list(geom_point = GeomPoint$desc), 
    examples = function(.) {
    })

geom_Rlogo <- function(mapping = NULL, data = NULL, stat = "identity", 
    position = "identity", na.rm = FALSE, show.legend = NA, inherit.aes = TRUE, 
    ...) {
    layer(data = data, mapping = mapping, stat = stat, geom = GeomRlogo, 
        position = position, show.legend = show.legend, inherit.aes = inherit.aes, 
        params = list(na.rm = na.rm, img = img, ...))
}
ggplot(mtcars, aes(wt, mpg))+geom_Rlogo()