Re: Idempotence and "Replication Insensitivity" are equivalent ?

From: vc <boston103_at_hotmail.com>
Date: 25 Sep 2006 12:06:10 -0700
Message-ID: <1159211170.535179.152510_at_e3g2000cwe.googlegroups.com>


pamelafluente_at_libero.it wrote:
> vc ha scritto:
>
> > pamelafluente_at_libero.it wrote:
> > > Both are invariant wrt linear transformations .... What do you think ?
> >
> > I think that you may consider reading a book or two on statistics
> > analysis if you are really interested in this kind of stuff.
>
> :))
> what? Another one?
>
>
> This sounds to me like an unnecessary rude statement.

I am not sure why you consider advice to read a book on the subject that might interest you rude, but that's OK. Surely, you did not expect to learn statistics from trivial exchanges in the database group.

>
> I see 2 possibilities:
>
> 1. You are implying I made some wrong statistical statement

You did not make any statistical statement. You stated a simple probability theory fact that the average is invariant under linear transformation. Quantiles are invariant to monotonic transformations. So what ? As I said before, whether invariance is important is something that you have to decide in the context of your model. Received on Mon Sep 25 2006 - 21:06:10 CEST

Original text of this message