R: ggplot ylim doesn't work

问题:

I am trying to change the ylim in a plot. But as soon as I change it, the plot vanishes. I am not able to figure out the reason. Can someone help me to root cause the issue?

>df<-data.frame(Category=c("A", "B", "C", "D", "B", "A", "D", "C"), Setup=c("X", "Y"), Gain=c(1.4, 1.45, 1.43, 1.48, 1.33, 1.37, 1.04, 1.09))
>p<-ggplot( df, aes(Setup, Gain)) + ylim(0.9, 1.25) + geom_bar(stat="identity") + facet_wrap( "Category", ncol=2 )

When I remove the ylim(), the above code works but with ylim it doesn't work. Any idea what's the issue?

Thanks.

回答1:

Replace ylim(0.9, 1.25) with

coord_cartesian(ylim = c(0.9, 1.25))

The function ylim also influences which data points are used for plotting. In contrast, coord_cartesian will not change the underlying data.

The complete command:

ggplot(df, aes(Setup, Gain)) + 
  coord_cartesian(ylim = c(0.9, 1.25)) +
  geom_bar(stat = "identity") + 
  facet_wrap("Category", ncol = 2)

The ylim approach will result in the following data:

df[df$Gain >= 0.9 & df$Gain <= 1.25, ]

  Category Setup Gain
7        D     X 1.04
8        C     Y 1.09

Furthermore, it will remove all bars since they start at 0.