以相同的比例和轴聚合和绘制直方图

Question

I have data named PACKS like this:

error    CENTRE_B
    -13  1104
    -13  1303
    -13  1303
    2    1204
    2    1403
    2    1403
    2    1403
    2    1502
    3    1503

My aim is to compare distributions. I want to to plot histograms for errors for each value of CENTRE_B. The problem is that the histograms have to be in the same scale.

I tried this:

new.par <- par(mfrow=c(5, 5))
histograms = aggregate(error ~ CENTRE_B, PACKS, hist)

This plots histograms. However, I don't know how to deliver additional argument to hist in aggregate (breaks = c(-80,80)).

Another problem with this is that histograms seems to be data.table of dimension 2x45, so it does not contain histograms. So I don't know how to make changing parameters automatic.

Thanks in advance for any suggestions.

Answer 1

Forget about aggregate . Just use split and do a simple for loop. Then you can input all args right into hist .

new=split(PACKS$error,PACKS$CENTRE_B)
for(i in 1:length(new))
hist(new[[i]],main="Some fancy title", sub="don't forget you can use paste to change the title between loops")