Jambo Frigerio wrote:

Il giorno lunedì 6 marzo 2017 alle 20:52:04 UTC+1 Auric__ ha scritto:
boxallbk wrote:

Hi Jerry,

If I want to do a linear regression with 1/x weighting, how do I
calculate the weights for the X data and Y data?

Thanks,

Baylye
You start by not replying to a post that's ***THIRTEEN*** ****ING YEARS
OLD!

****ing Google Groupies. You obviously have no idea what Usenet is or how it
works. In the future, when replying, remove everything you quote after the
"-- ", because that's my signature and has nothing to do with the body of
the post.

I am very sorry for your unfriendly answer to Baylye.

I'm not.

Weighted linear regression is still a widely used approach in analytical
chemistry and maybe in other fields.

No matter if the post is 13, or 50, or 100 years old!

Baylye's question was very interesting and an answer to it would benefit
other interested people.

He should have started a new thread, and referenced the previous thread.
Bringing an old thread back from the dead -- like you have done -- is not
The Right Thing[tm].

Google Groups is connected to Usenet, which predates GG by a couple DECADES.
The original users eventually came up with rules of etiquette ("netiquette")
that help keep the peace. GG users seem to be deliberately ignorant.

--
In my country, "lost in the wilderness" is the #2 leading cause of death.
Chipmunk attack is #4.