ggplotly unable to handle multiple legends properly in layered charts generated by ggplot2

[mjmg]

From the example in http://bxhorn.com/legends-ggplot/#managing-multiple-legends-in-a-layered-chart

# Multiple Legends
# load data
data(diamonds)
# calc quantiles with specified probabilities
new.prob <- c(0.90, 0.95, 0.99)
cut.probs <- plyr::ddply(diamonds, .(cut), numcolwise(quantile, probs = new.prob))[, c("cut", "price")]
cut.probs <- transform(cut.probs, quantile = c("P90", "P95", "P99"))
# layered graph
gg<-ggplot(diamonds, aes(cut, price)) + 
  geom_boxplot(aes(fill = factor(cut))) + 
  geom_point(data = cut.probs, aes(cut, price, color = factor(quantile)), size = 5) +
  scale_fill_discrete(name = "Quality of the Cut") +
  scale_color_discrete(name = "My Quantiles")
gg
plotly::ggplotly(gg)

1. Multiple legend titles are clustered together
2. Multiple legends are not separated
3. There is extra text ",1" in legend labels

This is the ggplot2 output:

This is the ggplotly output:

[cpsievert]

Will have to wait until plotly.js can support multiple legends -- plotly/plotly.js#1668

[henningsway]

Is there a way to avoid/ suppress the brackets, when combining the legends of two dimensions?

In my usecase it would be possible to paste two dimensions together e.g. fill and color.
fill would be "positive" or "negative" and the color would be some factorlabels: e.g. pos factor1 instead of (pos, factor1)?

[cgpu]

+1 Would also be interested in this feature. Even having the option to choose which one I want to choose would be a nicer option than the merge.

[jessicajturner]

Do you know if this has been fixed? I am still getting this error. If this has not been fixed, is there a way to work around it by using something other than ggplot or plotly?

[WhyBeeYes]

I think this issue is addressed in #572. The solution is to use aes outside geom for unwanted legends if the program is run in R Studio. In R Shiny, use style() function on the plotly object, as using the aes outside merges the legends together.