Better error message for plots (or facets with free scales) with all missing values.

[jasonelaw]

This bit me in a large,faceted plot where it wasn't entirely obvious what was going on:

d <- data.frame(x = 1:5, y = as.double(NA))
ggplot(data = d, aes(x, y)) + geom_point()
Error in seq.default(from = best$lmin, to = best$lmax, by = best$lstep) : 
  'from' must be of length 1

This is more like what caught me:

d <- expand.grid(x = 1:5, f = letters[1:5])
d$y <- rnorm(25)
is.na(d$y) <- d$f == 'a'
ggplot(data = d, aes(x, y)) + geom_point() + facet_wrap(~ f, scales = 'free_y')

Is it possible to catch this error?

I've encountered the same issue in a large dataset with faceted plots after melting the dataframe. Is there any workaround?

[hadley]

This sounds like a great feature/horrible bug, but unfortunately we don't currently have the development bandwidth to support it/fix it. If you'd like to submit a pull request that implements this feature/fixes this bug, please follow the instructions in the development vignette.

[johnaponte]

I have the same problem if the y values are Inf

[drolejoel]

not sure if this helps but there is a workaround with ggplot2 v1.0.1 (see http://stackoverflow.com/questions/23593159): The solution for me was to make sure my data.frame was nice and clean, by removing all NA values from the column with the Y axis data.

[Inventitech]

I just ran into this problem having upgraded to ggplot 2.0 (which was unrelated, but I did not know that at that time) and it took me 15 mins to figure out what was actually wrong. Yeah, if you can, please provide a better error message that makes it easier to spot that, for example, all your input data is NA.

[mattbk]

all your input data is NA

Not just for x and y, but other aesthetics as well.

[hadley]

No one supplied any reprexes apart from the initial, but I think this is fixed now.