Plotting multiple histograms quickly in R

Question

For exploratory analysis, its often useful to quickly plot multiple variables in one grid. An easy way to do this is to:

data(mtcars)    
hist(mtcars[,c(1,2,3,4)])

However, it becomes difficult to adjust breaks and axes to maintain consistency i.e.

hist(mtcars[,c(1,2,3,4)], breaks = 10)

does not effect the histograms. Is there an easy work around this or an easy way to do this in ggplot2?

Solution 1

This is how to do it with hist() :

lapply(mtcars[1:4], FUN=hist)

However I prefer to store plots in R objects with ggplot2 and display plot lists with cowplot::plotgrid() :

list <-lapply(1:ncol(mtcars),
              function(col) ggplot2::qplot(mtcars[[col]],
                                           geom = "histogram",
                                           binwidth = 1))

cowplot::plot_grid(plotlist = list)

Solution 2

With ggplot2 you can use facet_wrap to create a grid based on other variables.

For example:

library(ggplot2)

data(mtcars)

ggplot(data = mtcars) +
    geom_histogram(aes(x = mpg), bins = 4, colour = "black", fill = "white") +
    facet_wrap(~ gear)

And you can use the bins parameter to easily set how many breaks you want.

Comments

[amatsuo_net]

hist(mtcars[c(1,2,3,4)]) doesn't work. what do you see on your screen?

[abichat]

After loading mtcars, your code hist(mtcars[c(1,2,3,4),1]) produces an error: Error in hist.default(mtcars[c(1, 2, 3, 4)]): 'x' must be numeric

[Gooze]

Sorry for the late reply, i had computer issues and then forgot about this. The code runs fine for me in RStudio, however perhaps its best to put a comma before the column specification: hist(mtcars[,c(1,2,3,4)])

[Gregor Thomas]

This isn't quite what OP wants--OP is asking for how to plot histograms of each column, not of one column by another.

[Paolo]

@Gregor I realize that now after seeing the other answer (which works and seems to answer the question)

[Gooze]

Thanks, i think i can work with this by changing 1:ncol(mtcars) to 1:4. When i run lapply(mtcars[1:4], FUN=hist), however, i do not get the desired result above. Do you know why that is?