Hi, I am newby to R. I run this test (shapiro.test(rt(2000,30))) (R : Copyright 2001, The R Development Core Team Version 1.3.1 (2001-08-31)) for several times and results were different. Why do the shapiro.test produce different outputs? Thanks, -- Dr. Janos Korponai West-Transdanubian District Water Authority Dept. Kis-Balaton Csik F. str. 1 H-8360 Keszthely Hungary e-mail: korponai at georgikon.hu
On Monday 17 February 2003 18:57, Korponai J?nos wrote:> Hi, > > I am newby to R. I run this test (shapiro.test(rt(2000,30))) (R : > Copyright 2001, The R Development Core Team Version 1.3.1 > (2001-08-31))That version is ancient...by now there is R 1.6.2> for several times and results were different. > Why do the shapiro.test produce different outputs?If you generate random numbers by rt() you would expect the output to be different in each run, wouldn't you? If you generate the random numbers first: R> x <- rt(2000,30) and then run the test R> shapiro.test(x) it will indeed produce stable results. Z> Thanks,
Korponai J?nos <korponai at georgikon.hu> writes:> Hi, > > I am newby to R. I run this test (shapiro.test(rt(2000,30))) (R : Copyright > 2001, The R Development Core Team Version 1.3.1 (2001-08-31)) for several > times and results were different. > Why do the shapiro.test produce different outputs?Because you are calling it with different random samples from the t distribution! Look at help(rt), or just do rt(10,30) a couple of times. BTW, you should get a newer version. 1.6.2 is current. -- O__ ---- Peter Dalgaard Blegdamsvej 3 c/ /'_ --- Dept. of Biostatistics 2200 Cph. N (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk) FAX: (+45) 35327907