li - fit a baseline to stack 1 and remove it from the data or replace
	the data with it						RWW

gt,80a1300;us,80_160;li,1	Remove a linear baseline from 80a1300
				excluding chans 80 to 160 from the fit.
gt 80a1300;li,1,80		Remove a linear baseline from 80a1300
				excluding 80 chans around the center chan.
gt 80a1165;li,15,see:		Fit a 15'th order polynomial to scan 1165
				and replace stack 1 with the polynomial.

gt s3b4106 nbe:;st 2;gt bbe:;us -20_120;li 1 xt:;rt 1;pl
				Get the narrow and broad backend spectra,
				remove a linear baseline from the outer
				parts of the broad backend and use the same
				baseline on the narrow backend.  Plot it.

li makes a least squares fit of a polynomial of specified order (up to
15) to the data in stack 1. Channels for which the use array is 0 and
chans within wid of the sig or ref chan are excluded from the fit.

As indicated by the action node, the resulting polynomial will be
subtracted from stack 1 unless you request see. See causes stack 1 to
be replaced by the polynomial. "li xt:" causes the same polynomial which is
removed from stack 1 to be removed from stack 2 after being scaled. The
headers will be checked for compatibility.  If either stack 1 or 2 has
a baseline removed, the header variable rmval will be changed by the
amount of change in the expected signal channel (.expch1 or .expch2).

li does NOT recalculate the stack rms after the baseline removal.
The combination "li; rm; st ..." will store the new rms value in
the stack header variable .rms.

The value subtracted from the central channel is stored in the stack header
variable .rmval.

This document was last updated on Tue Oct 17 16:24:05 EDT 2000 by Marc W. Pound.