Standard functions for use with nl                              [STB-11: sg1.4]
----------------------------------

	^nl^ fcn depvar [varlist] ...                (see ^help nl^)

Standard fcns provided here are:

Exponential regression (one asymptote):
	^exp3^    Y = b0 + b1*b2^^X
	^exp2^    Y =      b1*b2^^X                   (i.e., b0 = 0)
	^exp2a^   Y =      b1*(1-b2^^X)

Logistic function (symmetric sigmoid shape):
	^log4^    Y = b0 + b1/(1+exp(-b2*(X-b3)))
	^log3^    Y =      b1/(1+exp(-b2*(X-b3)))     (i.e., b0 = 0)

Gompertz function (asymmetric sigmoid shape):
	^gom4^    Y = b0 + b1*exp(-exp(-b2(X-b3)))
	^gom3^    Y =      b1*exp(-exp(-b2(X-b3)))    (i.e., b0 = 0)



Description
-----------

An important feature of ^nl^, in addition to estimating arbitrary nonlinear
regressions, is the facility for adding prewritten common fcns.  Seven 
such functions are provided above.


Example
-------

^exp3^ is described above as providing:

	^exp3^    Y = b0 + b1*b2^^X

Typing 

	. ^nl exp3 ras dvl^

estimates a three-parameter exponential model (parameters b0, b1, b2) using
Y=ras and X=dvl.



Author
------

Patrick Royston, Royal Postgraduate Medical School, London.
	

Also see
--------

On-line:  ^help^ for ^nl^
    STB:  sg1.4 (STB-11)