how to add a plot to log plot in ggplot2?

Question

I ploting in log scale

library(ggplot2)
library(scales)
library(tibble)
tibble(
x = seq(1, 10, .01),
y = sin(log(x)*pi) + 2
) %>% 
ggplot(aes(x,y)) + 
geom_line() + 
scale_x_continuous(trans = log_trans()) + 
scale_y_continuous(trans = log_trans())

how can I add to this plot the data from

    tibble(
   x1 = seq(2, 11, .01),
   y1 = sin(log(x1)*pi) + 2
   )

typo, need to change log(x) to log(x1)

r2evans
– r2evans

2023-10-13 15:32:53 +00:00
Commented Oct 13, 2023 at 15:32 — r2evans
– r2evans, Commented Oct 13, 2023 at 15:32

r2evans · Accepted Answer · 2023-10-13 18:14:28Z

3

Two things:

you need to un-log transform the y1 data, since ggplot assumes all data needs to be transformed; and
you need to map the new variable names

Starting with your original plot unchanged (but assigned to gg):

gg <- tibble(
    x = seq(1, 10, .01),
    y = sin(log(x)*pi) + 2
  ) %>% 
  ggplot(aes(x,y)) + 
  geom_line() + 
  scale_x_continuous(trans = log_trans()) + 
  scale_y_continuous(trans = log_trans())

# fixing log(x) -> log(x1)
newdata <- tibble(
  x1 = seq(2, 11, .01),
  y1 = sin(log(x1)*pi) + 2
)

We can plot with:

gg + 
  geom_line(aes(x = x1),
            data = mutate(newdata, y = exp(y1)), color = "red")

FYI, unless you need scale_*_continuous for other reasons, I think scale_*_log10() presents better looking default ticks.

gg <- tibble(
    x = seq(1, 10, .01),
    y = sin(log(x)*pi) + 2
  ) %>% 
  ggplot(aes(x,y)) + 
  geom_line() + 
  scale_x_log10() + 
  scale_y_log10()
gg + 
  geom_line(aes(x = x1), data = mutate(newdata, y = exp(y1)), color = "red")

Another alternative (not sure what you're going for) is more generic and completely ancillary to the log-axis issue: we assign data=, but because the original aesthetics x and y are not both present, we need to map them explicitly.

gg +
  geom_line(aes(x = x1, y = y1), data = newdata, color = "red")

edited Oct 13, 2023 at 18:14

answered Oct 13, 2023 at 15:37

r2evans

167k8 gold badges92 silver badges176 bronze badges

Sign up to request clarification or add additional context in comments.

8 Comments

Tpellirn Over a year ago

thanks, but i think there is a problem in y scale, when we plot newdata alone, they are between 1 and 3. but when plotting it with gg they are now from 3 to above 10. any explainations?

r2evans Over a year ago

you tell me ... what exactly do you expect this to look like? You have non-log data that you plot in a lot scale, and then you have log-data ... do you expect it to be double-logged?

r2evans Over a year ago

see my edit @Tpellirn, perhaps that's it?

Tpellirn Over a year ago

in fact i am trying to plot like loglog in matlab

r2evans Over a year ago

That doesn't clear it up for me. I look at mathworks.com/help/matlab/ref/loglog.html and I see two scales, both in "log", with appropriate major/minor gridlines and a curve. That's what I think I'm giving you here. Please explain in words what is wrong with the curves we see here.

|

Collectives™ on Stack Overflow

how to add a plot to log plot in ggplot2?

1 Answer 1

8 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

8 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related