av.frame: Create data frame for added variable plot

In gmonette/spida: Collection of miscellaneous functions for mixed models etc. prepared for SPIDA 2009+ (development version)

Description

av.frame( model, variable) returns a data frame with model.frame(model) augmented by y.res and x.res, the residuals for an added variable plot

The purpose of this function is to facilitate OLS av.plots for mixed models.

Usage

av.frame(model, ...)

Arguments

model
...

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.
       library(nlme)
       library(lattice)
       hs <- read.csv( 'http://www.math.yorku.ca/~georges/Data/hs.csv')

       # Mixed model where ses and Sex are Level 1 and Sector is Level 2

       fit.mm <- lme( mathach ~ ses * Sex * Sector, hs, random = ~ 1+ses| school)

       # for diagnostics fit an OLS model using only level 1 variables interacting
       # with the id variable

       fit.ols <- lm( mathach ~ (ses * Sex ) * factor(school), hs)
       xyplot( y.res ~ x.res | factor(school), cbind(av.frame(fit.ols, 'ses:Sex'),hs), sub = 'ses:Sex')
       xyplot( y.res ~ x.res | factor(school), cbind(av.frame(fit.ols, '^Sex'),hs), sub = 'Sex')
       xyplot( y.res ~ x.res | factor(school), cbind(av.frame(fit.ols, '^ses$|^ses:f'),hs), sub = 'ses')


       Note : y.res is the residual from fitting the response on
              the model matrix for fit.ols omitting any column
              whose names is matched (as a regular expression)
              by 'effect'
              x.res is the residual of the first column of the
              model matrix that is matched by 'effect' on the
              same matrix used for y.res.
       Caution: To make sure that the correct columns were
              matched, the list of matched columns that are omitted
              is printed.

## The function is currently defined as
function (model, ...) 
{
    UseMethod("av.frame")
  }

gmonette/spida documentation built on May 17, 2019, 7:25 a.m.