R和ggplot2：为重叠范围创建摘要统计信息

Question

I have a set of records characterized by multiple variables, marked as either "in" or "out". I want to plot summary statistics for all records together and for those marked "in", while plotting each point only once, colored to show which ones are "in" or "out". How can I do that? I only know how to plot the summary statistics for the "in" and "out" groups (see code below), not for "in" and "all". It would be a plus if the legend explained the colors for the points (as in my illustration) as well as the colors for the error bars.

library(data.table)
library(ggplot2)
d = data.table(v1 = rnorm(10, 0, 1),
              v2 = rnorm(10, 1, 2),
              g = as.factor(c(rep('in', 7), rep('out', 3))))
m = melt(d, c('g'))
print(ggplot(m, aes(x = variable, y = value, colour = g)) +
      facet_wrap(~variable, scales = "free") +
      geom_jitter(position = position_jitter(height = 0, width = 0.2)) +
      stat_summary(fun.data = mean_se, geom = "errorbar", width = 0.25))

Answer 1

If you want to show in and out points, but errorbars for in and total, you should move your colour command and add different stat_summary for in and all:

library(data.table)
library(reshape2) #needed because data.table::melt will only work with reshape2
library(ggplot2)
d <- data.table(v1 = rnorm(10, 0, 1),
               v2 = rnorm(10, 1, 2),
               g = as.factor(c(rep('in', 7), rep('out', 3))))

m <- melt(d, c('g'))

ggplot(m, aes(x = variable, y = value)) + # removed colour here
        facet_wrap(~variable, scales = "free") +
        geom_jitter(aes(colour = g), position = position_jitter(height = 0, width = 0.2)) + #added color here
        stat_summary(fun.data = mean_se, geom = "errorbar", width = 0.25) + #errorbars for total observations
        stat_summary(data=m[m$g == "in",], fun.data = mean_se, geom = "errorbar", width = 0.25, colour = 2) # errorbars for "in" group