r - 如何摆脱 R 中 geom_point 中填充较深颜色的内点？

Question

我不希望图形中几何点中的那些更小、更内部、更深的彩色点。我该如何摆脱它？我试着玩弄形状，但什么也没发生。

library(stats)
library(ggfortify)
sp1 <- rnorm(72, mean = 4, 1)
sp2 <- rnorm(72, mean = 2, 1)
sp3 <- rnorm(72, mean = 3, 1)
sp4 <- rnorm(72, mean = 9, 1)
temp <- rnorm(72, mean = 20, 5)
season <- rep(c("Summer", "Autumn","Winter", "Spring"), each = 18)

sp.abd <- data.frame(sp1, sp2, sp3, sp4, season, temp)

pca_res <- prcomp(sp.abd[,-c(5,6)], scale. = TRUE)

autoplot(pca_res,
         loadings = TRUE, loadings.colour = 'blue',
         loadings.label = TRUE, loadings.label.size = 3) + 
  theme_classic() + geom_point(aes(color = sp.abd$season), size = 6, alpha = 0.5) + 
  scale_color_manual(values = rainbow(12,s = 0.6, start = 0, end = 0.7, rev = TRUE))

按照下面兰迪的回答，我尝试复制它以按温度着色（此处为渐变）

autoplot(pca_res, sp.abd, color  = temp, 
         loadings = TRUE, loadings.colour = 'blue',
         loadings.label = TRUE, loadings.label.size = 3)  + 
  scale_colour_gradient2(low = "white", mid = "cyan", high = "dodgerblue4", na.value = NA)

它只生成一个灰色图。只有当我在 autoplot 命令之外调用颜色时，它才会使用 scale_colour_gradient2 选项。但这仍然是点

autoplot(pca_res, size = 6, alpha. = 0.5,
         loadings = TRUE, loadings.colour = 'blue',
         loadings.label = TRUE, loadings.label.size = 3)  + geom_point(aes(colour = sp.abd$temp)) +
  scale_colour_gradient2(low = "white", mid = "cyan", high = "dodgerblue4", na.value = NA)

标签： rggplot2colorspca

Answer 1

看起来autoplot.prcomp（由指定的对象的autoplot方法）已经无需调用即可处理点，因此您可以将颜色映射、点大小和点不透明度放入其中：prcompggfortifygeom_point()autoplot

library(stats)
library(ggfortify)
sp1 <- rnorm(72, mean = 4, 1)
sp2 <- rnorm(72, mean = 2, 1)
sp3 <- rnorm(72, mean = 3, 1)
sp4 <- rnorm(72, mean = 9, 1)
season <- rep(c("Summer", "Autumn","Winter", "Spring"), each = 18)
sp.abd <- data.frame(sp1, sp2, sp3, sp4, season)
pca_res <- prcomp(sp.abd[,-5], scale. = TRUE)

autoplot(pca_res,
         ################
         # Add this stuff
         data = sp.abd, 
         color = 'season',
         size=6, alpha=.5,
         ################
         loadings = TRUE, loadings.colour = 'blue',
         loadings.label = TRUE, loadings.label.size = 3) + 
  theme_classic() + 
  scale_color_manual(values = rainbow(12,s = 0.6, start = 0, end = 0.7, rev = TRUE))