ggplot: How to increase space between axis labels for categorical data?

Question

I love ggplot, but find it hard to customize some elements such as X axis labels and grid lines. The title of the question says it all, but here's a reproducible example to go with it:

Reproducible example

library(ggplot2)
library(dplyr)

# Make a dataset
set.seed(123)
x1 <- c('2015_46','2015_47','2015_48','2015_49'
        ,'2015_50','2015_51','2015_52','2016_01',
        '2016_02','2016_03')
y1 <- runif(10,0.0,1.0)
y2 <- runif(10,0.5,2.0)


# Make the dataset ggplot friendly
df_wide <- data.table(x1, y1, y2)
df_long <- melt(df_wide, id = 'x1')

# Plot it
p <- ggplot(df_long, aes(x=x1, 
                         y=value, 
                         group=variable, 
                         colour=variable )) + geom_line(size=1)
plot(p)

# Now, plot the same thing with the same lines and numbers,
# but with increased space between x-axis labels
# and / or space between x-axis grid lines.

Plot1

The plot looks like this, and doesn't look too bad in it's current form:

Plot2

The problem occurs when the dataset gets bigger, and the labels on the x-axis start overlapping each other like this:

What I've tried so far:

I've made several attempts using scale_x_discrete as suggested here, but I've had no luck so far. What really bugs me is that I saw some tutorial about these things a while back, but despite two days of intense googling I just can't find it. I'm going to update this section when I try new things. I'm looking forward to your suggestions!

`+theme(axis.text.x = element_text(angle=90, vjust=0.5, size=10))` should help a bit for reading. — Haboryme, Oct 26 '16 at 14:24
If these are years and weeknumbers, you might want to use `scale_date` instead. That will allow you tot set breaks every year, for example. — Axeman, Oct 26 '16 at 14:55
If you translate you x values to valid dates ggplot will graw a nice time-axis. Currently those are strings — Ott Toomet, Oct 29 '16 at 06:32

score 2 · Accepted Answer · answered Nov 01 '16 at 18:09

As mentioned above, assuming that x1 represents a year_day, ggplot provides sensible defaults for date scales.

First make x1 into a valid date format, then plot as you already did:

df_long$x1 <- strptime(as.character(df_long$x1), format="%Y_%j")

ggplot(df_long, aes(x=x1, y=value, group=variable, colour=variable)) +
    geom_line(size=1)

The plot looks a little odd because of the disconnected time series, but scales_x_date() provides an easy way to customize the axis: http://docs.ggplot2.org/current/scale_date.html

ggplot: How to increase space between axis labels for categorical data?

1 Answers1