Preserving Slope after Changing Axis Limits (ggplot)?-CodePudding

Here is some hypothetical data:

data <- tibble(x=seq(1:20),
               y=c(55, 50, 45, 40, 42.5, 38, 35, 30, 29.5, 27, 
                     25, 23, 20, 15, 14, 12.5, 11, 8, 7, 6))

I would like to plot it on a logarithmic scale as such:

plot <- ggplot(data, aes(x, y))  
  geom_point()  
  geom_line()  
  scale_y_continuous(trans='log10', 
                     limits = c(1,100))  
  scale_x_continuous(breaks = seq(1,20, by=1),
                     limits = c(1,20))  
  theme_classic()  
  theme(aspect.ratio = 2)

plot

I intentionally set the aspect ratio to 2 (i.e. 2:1 y:x) . I would like to essentially truncate the y-axis to a minimum of 5 and a maximum of 60, however I want to preserve the same scaling factors. I don't want the new plot to have a 2:1 aspect ratio, but rather for it to be a "cropped" version of the full plot with a 2:1 aspect ratio. In other words, I don't want the slope of the line to change. However, when I try to adjust the limits of the y-axis, like below, the aspect ratio of 2:1, rather than the scaling factor (slope), is preserved. Is there any way to get around this? Thanks.

plot2 <- plot   ylim(5, 60)


plot2

CodePudding user response：

You could manually adjust the plot's dimensions by specifying the height and width using the ggplot function to "crop" the plot while maintaining the scaling factors. For instance:

plot <- ggplot(data, aes(x, y))  
  geom_point()  
  geom_line()  
  scale_y_continuous(limits = c(5,60), trans='log10')  
  scale_x_continuous(limits = c(1,20), breaks = seq(1,20, by=1))  
  coord_cartesian(ylim = c(5, 60))  
  theme_classic()

plot

CodePudding user response：

If you wish to preserve the slope of the line, then instead of using aspect.ratio, you can use coord_fixed, changing only the y limits:

plot <- ggplot(data, aes(x, y))  
  geom_point()  
  geom_line()  
  scale_y_continuous(trans='log10', 
                     limits = c(1,100))  
  scale_x_continuous(breaks = seq(1,20, by=1),
                     limits = c(1,20))  
  theme_classic() 

plot   coord_fixed(10)

plot   coord_fixed(10, ylim = c(5, 60))

^{Created on 2023-01-30 with reprex v2.0.2}