Comparing Kernel Density Estimation plots

Question

I am actually a novice to R and stats.. Could something like this be done in R

Determining the density estimates of two samples ( 2 Vectors )..?? I have done this Using R and obtained 2 density curves for the 2 samples using kernel density estimation ..

Is there anyway to quantitatively compare how similar/Dissimilar the density estimates of 2 samples are..?

I am trying to find out which data sample exhibits has a similar distribution to a particular distribution..

I am using R Language... Can somebody please help..??

YOu have an answer that was very helpful. Please click on the V-sign on the left of that answer to accept it, and do the same with the answers on your previous questions. This tells other people passing by that it was a correct answer. People will be less and less helpful when you don't acknowledge their answers. — Joris Meys, Apr 18 '11 at 11:22

score 7 · Accepted Answer · answered Apr 18 '11 at 06:55

7

You can use Kolmogorov-Smirnov test (ks.test) to compare two distributions. Cramer-von-Mises test is another one. There is this PDF Fitting Distributions with R where they also list other tests that are available (although the nortest package that he uses only tests for normality).

answered Apr 18 '11 at 06:55

Apprentice Queue

2,036
13
13

It seems that ks.test can be used only to compare to standard distributions such as Weibull, not comparing two non-standard – Herman Toothrot Aug 06 '21 at 09:33
@Toothrot, ks.test can test two arbitrary distributions; you optionally can refer to a named cumulative distribution family. – Apprentice Queue Nov 10 '21 at 16:02

score 4 · Answer 2 · answered Apr 18 '11 at 11:20

4

Apprentice Queue is right about using the Kolmogorov-Smirnoff test, but I wanted to add a warning: don't use it on its own. You should visually compare the distributions as well, either with two kernel density plots or histograms, or with a qqplot. Human brains are very good at playing spot-the-difference.

answered Apr 18 '11 at 11:20

Richie Cotton

118,240
47
247
360

Good point and, interestingly, that is exactly what @Pradeep proposes as his problem which neither answer currently offered really gets at. – Gavin Simpson Apr 18 '11 at 11:38
@Simpson, actually Pradeep's problem was he needed a quantitative procedure not a qualitative one. He had already calculated their densities and therefore could plot them. – Apprentice Queue Jul 04 '11 at 20:18

score 0 · Answer 3 · edited Oct 01 '12 at 15:51

0

You can try calculating the Earth mover's distance

edited Oct 01 '12 at 15:51

kaybenleroll

16,794
16
54
66

answered Jun 27 '11 at 21:06

skyde

2,816
4
34
53

Comparing Kernel Density Estimation plots

3 Answers3