Search code examples
rggplot2plot

gridArrange doesn't horizontally align anymore

Back in 2021 using gridArrange I was able to arrange two plots vertically so the values of the x-axis align perfectly. Here is the plot I produced then: well aligned plot

Notice how the x-axis values align in the top and bottom plot. Now I'm trying to produce a similar plot using new data. Here is the plot I produced: misaligned plot

In the present plot the plot areas are not aligned and so the x-axis values do not align vertically.

The code that produced these plots is virtually identical. Here is the 2021 code:

the.max <- ceiling(max(c(tab.13.3$diff, peeps$measure))) +1
the.min <- floor(min(c(tab.13.3$diff, peeps$measure))) -1

items.pts <- ggplot(tab.13.3, aes(x=diff, y=Item)) +
    geom_point() +
    scale_x_continuous(limits=c(the.min, the.max))

peeps.hist <- ggplot(peeps, aes(x=measure)) +
    geom_histogram() +
    scale_x_continuous(limits=c(the.min, the.max))


ggpeeps <- ggplotGrob(peeps.hist)
ggitems <- ggplotGrob(items.pts)

maxWidth = grid::unit.pmax(ggpeeps$widths[2:5], ggitems$widths[2:5])
ggpeeps$widths[2:5] <- as.list(maxWidth)
ggitems$widths[2:5] <- as.list(maxWidth)

grid.arrange(ggpeeps, ggitems, ncol=1,
                           heights=c(3, 2),
                           top=paste("Persons and Items for", params$themeas))

The data sets look like this: peeps:

  entry measure NAME      person.id network_id net.role
  <dbl>   <dbl> <chr>         <dbl>      <dbl>    <dbl>
1     1    4.86 4265  1 4      4265          1        4
2     2   -0.25 6353  1 4      6353          1        4
3     3    4.86 4273  1 4      4273          1        4
4     4    3.49 4256  1 4      4256          1        4
5     5    3.49 2545  1 3      2545          1        3
6     6    4.31 6352  1 4      6352          1        4

Items:

  Entry Item            name  value
  <dbl> <fct>           <chr> <dbl>
1     6 so_part_value_8 cat2  -0.95
2     6 so_part_value_8 cat3   0.37
3     6 so_part_value_8 cat4   1.76
4     6 so_part_value_8 cat5   4.38
5     1 so_part_ben_1   cat2  -3.89
6     1 so_part_ben_1   cat3  -1.42

The code I'm using now that results in misaligned axis values is:

the.max <- ceiling(max(c(tab.13.3$value, peeps$measure))) +1
the.min <- floor(min(c(tab.13.3$value, peeps$measure))) -1

items.pts <- ggplot(tab.13.3, aes(x=value, y=Item)) +
    geom_point() +
    scale_x_continuous(limits=c(the.min, the.max))

peeps.hist <- ggplot(peeps, aes(x=measure)) +
    geom_histogram() +
    scale_x_continuous(limits=c(the.min, the.max))


ggpeeps <- ggplotGrob(peeps.hist)
ggitems <- ggplotGrob(items.pts)

maxWidth = grid::unit.pmax(ggpeeps$widths[2:5], ggitems$widths[2:5])
ggpeeps$widths[2:5] <- as.list(maxWidth)
ggitems$widths[2:5] <- as.list(maxWidth)

grid.arrange(ggpeeps, ggitems, ncol=1, 
             heights=c(3, 2),
             top=paste("Persons and Items for", params$themeas))

The only thing I can think of is the gridArrange code changed. Can anyone tell me how I can make the present plot look like the 2021 version?

Many thanks.

Solution

  • Try using the patchwork package instead. It automatically align axis among other cool features.

    library(patchwork)
patchwork::wrap_plots(
  ncol=1,
  heights = c(3,2),
  peeps.hist,
  items.pts
)