[EMAIL PROTECTED] wrote:
> > Version 3.4 Release 1 for Sun SPARC, SunOS 5.3 : 1996
> >
> This is the version I have, and I just tried ks.gos on rnorm(100) twice,
> rnorm(1000) and rexp(1000) once each. In all 4 cases Splus rejected the
> the hypothesis that there's a fit at at least one sample p
In article <[EMAIL PROTECTED]>,
Cyril Goutte <[EMAIL PROTECTED]> wrote:
>
> Thanks for your reply.
> When I do ks.gof(1:10,rep(10,10)) on my system I get the same
> erroneous output I described earlier. The version I use is
> Version 3.4 Release 1 for Sun SPARC, SunOS 5.3 : 1996
> so hopefully t
Thanks for your reply.
When I do ks.gof(1:10,rep(10,10)) on my system I get the same
erroneous output I described earlier. The version I use is
Version 3.4 Release 1 for Sun SPARC, SunOS 5.3 : 1996
so hopefully this has been corrected on newer versions...
Best regards,
Cyril.
Dave Lorenz w
Cyril,
This is what i got from Splus when I tried the KS test.
> ks.gof(1:10,rep(10,10))
Two-Sample Kolmogorov-Smirnov Test
data: 1:10 and rep(10, 10)
ks = 0.9, p-value = 0.0002
alternative hypothesis:
cdf of 1:10 does not equal the
cdf of rep(10, 10) for at least
Hi all,
I have a little worry with the way Splus calculates the two-sample
Kolmogorov-Smirnov statistics.
Consider the following samples:
---
> ds1
[1] 1 2 3 4 5 6 7 8 9 10
> ds2
[1] 10 10 10 10 10 10 10 10 10 10
---
The output of the KS test is:
---
> ks.gof(ds1,ds2)
Two