bupdate: Update the b fixed effect to minimise the b-c random effect...
In sitar: Super Imposition by Translation and Rotation Growth Curve Analysis

bupdate

R Documentation

Update the b fixed effect to minimise the b-c random effect correlation

Description

A function to update the value of bstart, the starting value for the b fixed effect, to minimise the correlation between the random effects b and c.

Usage

bupdate(x)

Arguments

`x`	a `sitar` object.

Value

Returns an updated value of the b fixed effect, based on the random effect covariance matrix.

Author(s)

Tim Cole tim.cole@ucl.ac.uk

Examples


## fit sitar model with b fixed effect starting value defaulting to 'mean'
m1 <- sitar(x=age, y=height, id=id, data=heights, df=5)
print(fixef(m1)['b'])

## refit with starting value chosen to minimise b-c correlation and df increased
m2 <- update(m1, bstart=bupdate(m1), df=6)
print(fixef(m2)['b'])

sitar documentation built on July 9, 2023, 6:51 p.m.