Rui,
Your response nearly answered a similar question of mine except that I also
have ecdfs of different lengths.
Do you know how I can adjust x - seq(min(loga, logb), max(loga, logb),
length.out=length(loga))
to account for this? It must be in length.out() but I'm unsure how to
proceed.
Hello,
Try length.out = max(length(loga), length(logb))
Note also that all of the previous code and the line above assumes that
we are interested in the max distance, whereas the KS statistic computes
the supremum of the distance. If it's a two sample test then their
values are almost surely
Subject:
Re: [R] Kolmogorov-Smirnov test and the plot of max distance between two
ecdf curves
Sent by:
r-help-boun...@r-project.org
Rui,
Your response nearly answered a similar question of mine except that I
also
have ecdfs of different lengths.
Do you know how I can adjust x - seq(min
thanks rui
that's what I was looking for
I have another related question:
- why of the difference between the max distance D calculated with ks.test()
and the max distance D “manually” calculated as in (2)?
I guess it has something to do with the fact that KS is obtained with a
maximisation
Hello,
That's a very difficult question. See
Marsaglia, Tsang, Wang (2003)
http://www.jstatsoft.org/v08/i18/
Simard, L'Ecuyer (2011)
http://www.jstatsoft.org/v39/i11
R's ks functions are a port of Marsaglia et al. to the .C interface.
Rui Barradas
maxbre wrote
thanks rui
that's what I
thanks for the help: I'll have a look at the papers
max
Il 28/05/2012 12:31, Rui Barradas [via R] ha scritto:
Hello,
That's a very difficult question. See
Marsaglia, Tsang, Wang (2003)
http://www.jstatsoft.org/v08/i18/
Simard, L'Ecuyer (2011)
http://www.jstatsoft.org/v39/i11
R's ks
Just a final correction.
I was wrong, stats::ks.test doesn't use only Marsaglia et al.
It's even clearly written in the help page.
Read the documentation before stating!
Rui Barradas
Em 28-05-2012 11:51, maxbre escreveu:
thanks for the help: I'll have a look at the papers
max
Il 28/05/2012
Hello,
Try the following.
(i've changed the color of the first ecdf.)
loga - log10(a+1) # do this
logb - log10(b+1) # only once
f.a - ecdf(loga)
f.b - ecdf(logb)
# (2) max distance D
x - seq(min(loga, logb), max(loga, logb), length.out=length(loga))
x0 - x[which( abs(f.a(x) - f.b(x)) ==
Hi all,
given this example
#start
a-c(0,70,50,100,70,650,1300,6900,1780,4930,1120,700,190,940,
760,100,300,36270,5610,249680,1760,4040,164890,17230,75140,1870,22380,5890,2430)
length(a)
b-c(0,0,10,30,50,440,1000,140,70,90,60,60,20,90,180,30,90,
9 matches
Mail list logo