Re: correlation

What do you mean by the " r square value" ?
You must be clear.
Why don't you square the coefficient?

r2 = ms^2

On 3/18/11 3:51 PM, mark vogel wrote:
> I have a tquestion about correlation too I wonder what is the best way
> to calculate r square value
> iv = 4753
> x = v11(10753:14352:2)
> y = v2tem(3216:5015)
> print(x)
> print(y)
> ms = regline(x,y)
> rc = regCoef(x,y)
> data1 = new ((/2,1800/),float)
> data1(0,:) = y
> data1(1,:) = ms*(x-ms@xave <mailto:x-ms@xave>) + ms@yave <mailto:ms@yave>
> corr = esccr(x,y,0)
>
> Is it correct?
> On Fri, Mar 18, 2011 at 12:54 PM, Dennis Shea <shea@ucar.edu
> <mailto:shea@ucar.edu>> wrote:
>
> The reason they were removed was that there was an error.
> I wrote a function in NCL and will send to you offline.
>
> Regards
> D
>
> On 3/17/11 4:38 PM, Rob Nicholas wrote:
> > The pattern correlation examples mentioned in the middle of the page
> > for escorc() and in various posts to this list seem to have
> > disappeared. I haven't been able to find them elsewhere on the NCL
> > site -- do they still exist?
> >
> > ~Rob
> > _______________________________________________
> > ncl-talk mailing list
> > List instructions, subscriber options, unsubscribe:
> > http://mailman.ucar.edu/mailman/listinfo/ncl-talk
> _______________________________________________
> ncl-talk mailing list
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>
>
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Sun Mar 20 11:13:49 2011