richards-package {richards} | R Documentation |
selfStart models and functions for Richards curves:
SSrichardsSimple <- selfStart( ~ d + (a - d) / (1 + (2^ny-1) * (input / x50)^b)^(1/ny), function (mCall, data, LHS) initialRichards(mCall, data, LHS), c("a", "d", "b", "x50", "ny"))
Why a package for this one line?
The function for the initial values is a call to the selfStart model SSfpl of the four parameter logistic function.
selfStart models for Richards curves with specific fixed values
of the shape parameter ny
are also given.
For these a Richards function described by ny
and the same asymptotes, location and slope parameters as
for a fitted four parameter logistic function
is found and used for the initial value.
If ny
is negative then 1 + (2^ny-1)*(input/x50)^b
can
also be negative, and then this value cannot be raised to 1/ny
(unless 1/ny
is an integer).
'Modified' selfStart models are given to handle this situation.
Package: | richards |
Type: | Package |
Version: | 0.4.5 |
Date: | 2008-09-01 |
License: | GPL Version 2 or later |
Jens Henrik Badsberg
Maintainer: Jens Henrik Badsberg <september2008richards@badsberg.eu>
Richards, F.J. 1959: A flexible growth function for empirical use. J. Exp. Bot. 10: 290-300.
http://en.wikipedia.org/wiki/Generalised_logistic_curve.