# NOT RUN {
# Generate data: means and standard errors of means for prices
# for each type of cut
dmod <- lm(price ~ cut, data = diamonds)
cut <- unique(diamonds$cut)
cuts_df <- data.frame(
cut,
predict(dmod, data.frame(cut), se = TRUE)[c("fit", "se.fit")]
)
ggplot(cuts_df) +
aes(
x = cut,
y = fit,
ymin = fit - se.fit,
ymax = fit + se.fit,
colour = cut
) +
geom_pointrange()
# Using annotate
p <- ggplot(mtcars, aes(x = wt, y = mpg)) + geom_point()
p
p + annotate(
"rect", xmin = 2, xmax = 3.5, ymin = 2, ymax = 25,
fill = "dark grey", alpha = .5
)
# Geom_segment examples
p + geom_segment(
aes(x = 2, y = 15, xend = 2, yend = 25),
arrow = arrow(length = unit(0.5, "cm"))
)
p + geom_segment(
aes(x = 2, y = 15, xend = 3, yend = 15),
arrow = arrow(length = unit(0.5, "cm"))
)
p + geom_segment(
aes(x = 5, y = 30, xend = 3.5, yend = 25),
arrow = arrow(length = unit(0.5, "cm"))
)
# You can also use geom_segment() to recreate plot(type = "h")
# from base R:
counts <- as.data.frame(table(x = rpois(100, 5)))
counts$x <- as.numeric(as.character(counts$x))
with(counts, plot(x, Freq, type = "h", lwd = 10))
ggplot(counts, aes(x = x, y = Freq)) +
geom_segment(aes(yend = 0, xend = x), size = 10)
# }
Run the code above in your browser using DataLab