sinhrks/ggfortify

ggdistribution

Opened this issue · 3 comments

Deleted user commented

Why
ggdistribution(ppois, seq(0, 30), lambda = 20)
gives me continous function?

sinhrks commented

Thx for the report. It is caused by the line is drawned with geon line or ribbon. It should use step for discrete distributions.

ggfortify/R/fortify_stats_density.R

Line 53 in 55b60ce

p <- p + geom_factory(geom_ribbon, object, colour = colour,

PR is appreciated:)

lselzer commented

If it plots the probability of a discrete distribution then it should be points or a thin line. And I have an issue with the percentage in the Y axis with the pdf of continous distribution as it seems to imply that this is a probability.

Anyway, do you guys think it would be possible to infer if its a discrete distribution otherwise than checking the call?

terrytangyuan commented

@lselzer I think we'll have to check the call to infer whether it's discrete or not since results from stats does not give that information to us directly.