Question

I have two different datasets (c1 and c2), which are plotted together in one graph. both curves have different x- and y-values:

c1 = data.frame(
  x=c(0,1.1,2,  3,  4,  5),
  y=c(0,1.1,1.9,3.2,4.3,5.2)
)
c2 = data.frame(
  x=c(0,0.3,0.9,2.1,3.2,4.2,5),
  y=c(0,0.4,1.5,2.3,3.2,4.1,5.1)
)
plot(c1, type="o", col=2)
lines(c2, type="o", col=3)

Now I like to plot the residuals of the two curves (res=c1-c2) for all unique x values (unique(c(c1$x, c2$x))). This would be easy if I had the same x-values. But it seems, that I have to interpolate all missing x-values and add them to the measured dataset.

Is there an easy way for doing this in R?

Answer 1

How about this:

On the values c1$x:

c1$y - approx(c2$x, c2$y, c1$x)$y
# [1]  0.0000000 -0.5333333 -0.3333333  0.1636364  0.3800000  0.1000000

on the values c2$x:

approx(c1$x, c1$y, c2$x)$y - c2$y
# [1]  0.00 -0.10 -0.60 -0.27  0.22  0.38  0.10

Or, putting it all together,

x <- sort(unique(c(c1$x, c2$x)))
approx(c1$x, c1$y, x)$y - approx(c2$x, c2$y, x)$y
# [1] 0.0000000 -0.1000000 -0.6000000 -0.5333333 -0.3333333 -0.2700000  0.1636364  0.2200000
# [9] 0.3800000  0.3800000  0.1000000