Linest question

Arokh · Newbie Joined: 05 Sep 2007 Posts: 2

I've tried using the linest() function on a set of only two values. By forcing the regression to having no intercept, I thought I had "three" data points. The linest() function however doesn't compute any error for the slope and the R² is 1.00. The slope itself is computed correctly.

Is this behaviour by design or did I use a wrong parameter?

the data array is:
0.1 0.67464
0.2 0.79617

David · Posted: Wed Sep 05, 2007 9:44 am Post subject: Re: Linest question

Why would you do that? Isn't that cracking an egg with a sledgehammer? The manual calculations are dead simple.

Arokh · Newbie Joined: 05 Sep 2007 Posts: 2

I know that the calulations are fairly simple. it's just proof of principle. I thought the idea of using calc is not having to do any manual calculations anymore. Smile

I'm a convert from Excel/Origin to Calc and in this particular example the regression worked in those two programs but it didn't work in Calc.

After experimentig a bit I found that adding a third "real" data point (0/0) made the linest-function compute the stats correctly. *confusing*

David · Posted: Wed Sep 05, 2007 6:44 pm Post subject:

No. That's the same excuse given for the calculator, and both have successfully produced one numerically illiterate nation after another. The computer is useful for making easier onerous, tedious, repetitive, or extremely complex calculations. It is not intended, nor should it be used for mundane, simple-minded tasks.

Sorry ...ex- high school math teacher here ranting about the state of the world ...but I know I'm right on that.

David.