How do I remove a gap in date x axis?

Question

I have a time series data with a gap of year (e.g. 2006). I attempted to create a continuous lineplot ignoring this gap with scale_x_date. However, the function only removes the date labels on the x-axis and leaves a gap in the plot. Is there a way to remove the x-axis text and the gap in datapoints?

Data:

set.seed(123)
n <- 100
y <- sample.int(n, replace=TRUE)
t <- seq(from=as.Date("2001-01-01"), by="month", length.out=n)
dat <- data.frame(t, y)[!format(t, "%Y") %in% 2006, ]

Tried:

library(ggplot2)
my_break_fun <- function(x){
  br <- seq(from = min(x), to=max(x), by="2 month")
  br[!format(br, "%Y") %in% 2006]
}

dat |> 
  ggplot(aes(x=t, y=y)) +
  geom_line() +
  scale_x_date(
    date_labels="%b-%Y",
    breaks=my_break_fun
  ) +
  theme(axis.text.x = element_text(angle=90, hjust=1, vjust=1))

Answer 1

Facets would be one common way:

dat |> 
  dplyr::mutate(grp = t >= as.Date("2006-01-01")) |>
  ggplot(aes(x=t, y=y)) +
  geom_line() +
  scale_x_date(
    date_labels="%b-%Y",
    breaks=my_break_fun
  ) +
  facet_grid(.~grp, scales = "free_x", space = "free_x") +
  theme(axis.text.x = element_text(angle=90, hjust=1, vjust=1)) +
  theme(strip.background = element_blank(),
        strip.text = element_blank())

Answer 2

We can use {ggbreak}:

dat |> 
  ggplot(aes(x=t, y=y)) +
  geom_line() +
  ggbreak::scale_x_break(c(as.Date("2005-12-02"), as.Date("2006-12-31")), 
                         expand = FALSE) +
  scale_x_date(date_labels="%b-%Y", breaks="2 months") +
  theme(axis.text.x = element_text(angle=90, hjust=1, vjust=1),
        axis.text.x.top = element_blank(),
        axis.ticks.x.top = element_blank())

^{Created on 2025-09-16 with reprex v2.1.1}

Answer 3

Base R, addressing breaks by different line colours.

dat = data.frame(t, y)[
  !format(t, '%Y') %in% 2006, ] |> 
  transform(i=seq_along(t))

L = split(dat, cumsum(c(FALSE, diff(dat$t)>31)))
n = palette.colors(length(L))

par(mar=c(6, 4.1, 4.1, 2.1))
plot(y~i, data=dat, type='n', xaxt='n')
for (l in seq_along(L)) 
  lines(L[[l]]$i, L[[l]]$y, col=n[l], type='b')
axis(1, at=dat$i, labels=format(dat$t, '%b-%Y'), las=2)

---

Note

In my (publication) routine, when creating several figures in one script, I tend to wrap the figure-generating lines in local({ <code here> }), to prevent the global environment from getting cluttered.

Answer 4

I know this is tagged ggplot2, but a base R solution for others not using ggplot would be to create an index (here, seq_along(dat$t)) and then label the axis based on the index:

par(mar = c(6, 4.1, 4.1, 2.1)) # just to make room for long dates

plot(seq_along(dat$t), dat$y, type = "l", 
     xaxt = "n", xlab = "", ylab="Value")

axis(1, at = seq_along(dat$t), labels = labels = format(dat$t, "%b-%Y"), las = 2)

This is generally bad practice in the visual representation of data since it doesnt address the gap in data. Another way indicate the gap in data (other than changing line color) would be to add a vertical line with abline. This can also be done by simply creating an indicator using which and diff that "finds" the gap:

abline(v = which(as.numeric(diff(dat$t)) > median(diffs)) + 0.5,
       col = "red", lty = 3, lwd = 2)