Re: regressions?

[verschel@xxxxxxxxxxxxxxxxxxxxx (Mark A. Verschell)]

You bet, Ilana.

If you do a Fgo regress:
* * * * * * * * in /usr/local/ferret/go
regresst.jnl: define FERRET variables for regression along the T axis
regressx.jnl: define FERRET variables for regression along the X axis
regressy.jnl: define FERRET variables for regression along the Y axis
regressz.jnl: define FERRET variables for regression along the Z axis

use regresst (regression in time), thusly:

yes? let p=var2    !independent variable
yes? let q=var1    !dependent variable
yes? go regresst

You can examine the results either statistically:

yes? list slope,intercep,rsquare

or graphically (this can be done a lot of ways):

yes? plot/vs p,q                   ! the original scatter plot
yes? plot/vs/line=2/over p,qhat    ! the regression line in red

Mark

On Jun 15,  2:48pm, Ilana WAINER wrote:
> Subject: regressions?
> I would like to tap (once more) the knowledge of this group: is there any way
to
> obtain the regression line (and coefficients) when doing a scatter plot of
var1
> versus var2, i.e. var1 = A*var2+B?
>
> thanks again,
>
> Ilana Wainer
>-- End of excerpt from Ilana WAINER


-- 
 Mark Verschell                     NASA/Goddard Space Flight Center
 verschell@neptune.gsfc.nasa.gov    Code 970/Lab. for Hydrospheric Proc.
 301-286-2027  Fax: 301-286-1761    Greenbelt, MD 20771