ggplot2 scale_y_log10 不适用于堆叠的 geom_histogram

编程问答 2022-04-22

问题描述

structure(list(Hypothesis = c("H0","H0","H1","H1"),PP = c(9.26073693450588e-29,3.34281982011265e-24,9.38662550316043e-15,7.95890908425815e-06,2.86082492381069e-28,2.20405950836506e-13,9.0925912086661e-35,2.31002748754231e-42,4.76347929684044e-23,1.38383109879787e-23,2.86033450403371e-19,1.18097638779878e-220,1.10136930662603e-20,3.34955010022165e-17,1.12279490220994e-28,3.58934964238325e-21,4.82006911711899e-21,2.3176322957431e-273,3.65469210312312e-23,1.035293515573e-60,1.03909796125627e-46,1.72578715346419e-21,0.096735363503543,3.97631648427947e-16,0.00403790620239148,0.235149734366432,0.298832898846241,0.0948134291565298,0.160951055566603)),row.names = c(NA,30L),class = "data.frame")

我正在尝试使用 `:

绘制直方图 
p<-ggplot(t,aes(x=PP,fill=Hypothesis))+
geom_histogram(binwidth = 0.01)
p

enter image description here

如果我使用 scale_y_log10()，Y 轴上的值会变大而不是对数缩放

p<-ggplot(t,fill=Hypothesis))+
geom_histogram(binwidth = 0.01)+
scale_y_log10()

知道为什么 PP = 0 的计数会变高吗？

log scale

解决方法

我们可以使用position= 'dodge'

library(ggplot2)
p <- ggplot(t,aes(x = PP,fill = Hypothesis))+
   geom_histogram(binwidth = 0.01,position = 'dodge')+
   scale_y_log10()
ggplot2histogramlogarithmrrstacked-chart