clauswilke
commented
5 years ago For a dataset of this size, please make a reprex with computationally generated data rather than a gigantic dput dump.
I can't reproduce your crash with a similarly sized data set, but it's clear that you're misusing the ggplot2 API and that causes the slow rendering. You're drawing 10000 arrows and 10000 text labels on top of each other. Once that is fixed, rendering is reasonably fast.
library(ggplot2)
df <- data.frame(trans = 10*rnorm(10000))
ggplot(df, aes(x = trans)) +
geom_density() +
geom_vline(xintercept = mean(df$trans), col = "green", size =2) +
geom_curve(data = data.frame(x = 1), aes(x = 10, y = .15, xend = 4.4, yend = .18),
colour = "#555555", size=0.5, curvature = 0.3,
arrow = arrow(length = unit(0.03, "npc"))) +
geom_text(data = data.frame(x = 1), aes(x= 9, y = .15, label= paste0("Average number of items\n per transaction is: ",
round(mean(mtcars$mpg),2)), colour = "blue", family = "Times New Roman")) 
Created on 2019-08-27 by the reprex package (v0.3.0)
abiyug
yutannihilation
lock[bot]
I am working with a data that has 10 thousand observations and 2 vars. When attempting to add a geom_curve layer, the rendering becomes extremely slow to a point where it distorts the plot and eventually crashes and generate the output below.