ts
>>> (the posting guide is that clearly labeled bit
>>> at the bottom that looks like this:
>>> PLEASE do read the posting guide
>>> http://www.R-project.org/posting-guide.html
>>> Check it out! It is cool.)
>>>
>>> Additionally
he R maintainers do an amazing job of creating
> numerically stable platform-independent software,
> so you get the same results almost everywhere.
> (Thank you R core!)
>
>
> HTH
>
> Steve McKinney
>
> -Original Message-
> From: [EMAIL PROTECTED] on
self-contained
>> and reproducible set of R commands).
>>
>> Finally, describe what the difference is and why
>> the difference is problematic (i.e. don't report
>> machine precision differences, or sign differences
>> for PCA results - PCA vector directions are
able platform-independent software,
> so you get the same results almost everywhere.
> (Thank you R core!)
>
>
> HTH
>
> Steve McKinney
>
> -----Original Message-
> From: [EMAIL PROTECTED] on behalf of jathine
> Sent: Tue 9/16/2008 2:19 PM
> To: r-help@
Steve McKinney
-Original Message-
From: [EMAIL PROTECTED] on behalf of jathine
Sent: Tue 9/16/2008 2:19 PM
To: r-help@r-project.org
Subject: [R] Different PCA results under Windows and Linux
I ran the following R script under both Linux and Windows, and got 2
different results.
Linux R version 2.7.1
I ran the following R script under both Linux and Windows, and got 2
different results.
Linux R version 2.7.1 and Windows R version 2.7.2.
> library(FactoMineR)
>x1=read.table("freqtest.txt",header=TRUE)
>xrcc2=x1[,1:8]
>p1=PCA(xrcc2, graph=FALSE)
>p1$var
freqtest.txt file lines of text :
M1 M2
6 matches
Mail list logo
