CODEWITHSUNDEEP
rplotggplot2
VDK
VDK

Reputation: 819

Subgroup axes ggplot2 and axis limits

Follow up to:

Subgroup axes ggplot2 similar to Excel PivotChart

ggplot2 multiple sub groups of a bar chart

R version 3.1.1 (2014-07-10) Platform: i386-w64-mingw32/i386 (32-bit)

I am working on a plot with ggplot2. The aim is to tweak the axis into a look similar to Excels famous pivot graphs. I know, how I can achieve the look I want, but as soon as I use axis limits, the code is not sufficient any more.

Data:

library(reshape2)
library(ggplot2)
library(grid)

df=data.frame(year=rep(2010:2014,each=4),
          quarter=rep(c("Q1","Q2","Q3","Q4"),5),
          da=c(46,47,51,50,56.3,53.6,55.8,58.9,61.0,63,58.8,62.5,59.5,61.7,60.6,63.9,68.4,62.2,62,70.4))

df.m <- melt(data = df,id.vars = c("year","quarter"))

g1 <- ggplot(data = df.m, aes(x = interaction(quarter,year), y = value, group = variable)) +
      geom_area(fill = "red")+
      coord_cartesian(ylim = c(0, 75)) +
      annotate(geom = "text", x = seq_len(nrow(df)), y = -1.5, label = df$quarter, size = 2, color = "gray48") +
      annotate(geom = "text", x = 2.5 + 4 * (0:4), y = -3, label = unique(df$year), size = 3, color ="gray48") +
      theme_grey(base_size = 10)+
      theme(line = element_line(size = 0.2),
            axis.title.x = element_blank(),
            axis.text.x = element_blank(),
            legend.position= "none")

#remove clipping of x axis labels
g2 <- ggplot_gtable(ggplot_build(g1))
g2$layout$clip[g2$layout$name == "panel"] <- "off"
grid.draw(g2)

png(filename = "test.png",width = 14/2.54,height = 6/2.54, units = "in",res = 300)
grid.draw(g2)
dev.off()

enter image description here

The plot is fine and the axis lables are as wished. But as soon as you change the limits of the y axis everything is messed up.

enter image description here

I hope you have an idea, how to solve my problem!

Upvotes: 2

Views: 1134

Answers (1)

Sandy Muspratt
Sandy Muspratt

Reputation: 32789

Actually, it is plotting exactly what you are asking for. Check ?geom_area, and you will note that the minimum y is 0. So when you turn off clipping, ggplot will show as much of the area as it can within the limits of the lower margin. Instead use geom_ribbon(). It has ymax and ymin. Also, you need to take care setting the y-coordinates in the two annotate() functions.

library(reshape2)
library(ggplot2)
library(grid)

df=data.frame(year=rep(2010:2014,each=4),
   quarter=rep(c("Q1","Q2","Q3","Q4"),5),
   da=c(46,47,51,50,56.3,53.6,55.8,58.9,61.0,63,58.8,62.5,59.5,61.7,60.6,63.9,68.4,62.2,62,70.4))

df.m <- melt(data = df,id.vars = c("year","quarter"))

ymin <- 40 
g1 <- ggplot(data = df.m, aes(x = interaction(quarter,year), ymax = value, group = variable)) +
      geom_ribbon(aes(ymin=ymin), fill = "red")+
      coord_cartesian(ylim = c(ymin, 75)) +
      annotate(geom = "text", x = seq_len(nrow(df)), y = 37.5, label = df$quarter, size = 2, color = "gray48") +
      annotate(geom = "text", x = 2.5 + 4 * (0:4), y = 36.5, label = unique(df$year), size = 3, color ="gray48") +
      theme_grey(base_size = 10)+
      theme(line = element_line(size = 0.2),
            axis.title.x = element_blank(),
            axis.text.x = element_blank(),
            legend.position= "none",
            plot.margin = unit(c(1,1,3,1), "lines"))     # The bottom margin is exaggerated a little

# turn off clipping of the panel
g2 <- ggplotGrob(g1)
g2$layout$clip[g2$layout$name == "panel"] <- "off"
grid.draw(g2)

enter image description here

Upvotes: 1

Related Questions