Categorize data with geom_col

[guest777]

Hello, I am making a simple column chart, I use aes(fill = ) to plot 2 categories by date. And I also want to somehow highlight the first two columns (change color or add hatching?) associated with the first date (May 2). Is there an easy way to do this? Thanks!

[nirgrahamuk]

I think the package gghighight would allow this.
https://cran.r-project.org/web/packages/gghighlight/vignettes/gghighlight.html

[FJCC]

Here is a manual method using the color aesthetic.

library(ggplot2)
library(dplyr, warn.conflicts = FALSE)
DF <- data.frame(Date = rep(c("2019-05-02", "2019-05-28", "2019-06-11", "2019-08-10"), each = 2),
                 Value = c(33, 44, 25, 21, 35, 28, 33, 33),
                 Pop = rep(c("A", "B"), 4))
DF <- DF %>% mutate(Highlight = ifelse(Date == "2019-05-02", "May_2", "Not_May_2"))
Fill_COLORS = c("A" = "skyblue", "B" = "blue")
Colors = c("May_2" = "red", "Not_May_2" = "black")
ggplot(DF, aes(Date, Value, fill = Pop, color = Highlight)) + 
  geom_col(position = "dodge", size = 1.2) +
  scale_fill_manual(values = Fill_COLORS) +
  scale_color_manual(values = Colors) +
  guides(color = FALSE, fill = FALSE)

^{Created on 2020-07-12 by the reprex package (v0.3.0)}

[Z3tt]

In case you do not use gridlines you could also use transparency:

library(ggplot2)
library(dplyr, warn.conflicts = FALSE)
DF <- data.frame(Date = rep(c("2019-05-02", "2019-05-28", "2019-06-11", "2019-08-10"), each = 2),
                 Value = c(33, 44, 25, 21, 35, 28, 33, 33),
                 Pop = rep(c("A", "B"), 4))
DF <- DF %>% mutate(Highlight = ifelse(Date == "2019-05-02", "May_2", "Not_May_2"))

Fill_COLORS = c("A" = "skyblue", "B" = "blue")

ggplot(DF, aes(Date, Value, fill = Pop, alpha = Highlight)) + 
  geom_col(position = "dodge", size = 1.2) +
  scale_fill_manual(values = Fill_COLORS, name = "Pop") +
  scale_alpha_manual(values = c(1, .3), guide = F) +
  theme_minimal() +
  theme(panel.grid = element_blank())

^{Created on 2020-07-13 by the reprex package (v0.3.0)}