plotdhglm: Produce Model-Checking Plots for a Double Hierarchical...

In dhglm: Double Hierarchical Generalized Linear Models

Description

Plots residuals for the mean and dispersion models

Usage

plotdhglm(OUTPUT, type="mean", random=NULL)

Arguments

OUTPUT	The <dhglmfit> object to be plotted
type	Type of model required (mean, dispersion)
random	Random term whose residuals are to be plotted (mean, phi, v, alpha). Default (NULL) is the residuals from the full model

Details

Four types of plot are available (normal probability plot for residuals, histogram of residuals, residuals versus fitted values and absolute values of residuals versus fitted values).

See Also

<dhglmfit>

Examples

#### Model checking plot for crack-growth data
data(crack_growth)
model_mu<-DHGLMMODELING(Model="mean", Link="log", 
LinPred=y~crack0+(1|specimen),RandDist="inverse-gamma")
model_phi<-DHGLMMODELING(Model="dispersion", Link="log", 
LinPred=phi~cycle+(1|specimen),RandDist="gaussian")
res_crack_dhglm1<-dhglmfit(RespDist="gamma", DataMain=crack_growth, 
MeanModel=model_mu, DispersionModel=model_phi,Maxiter=1)
plotdhglm(res_crack_dhglm1)

Example output

Loading required package: Matrix
Loading required package: boot
Loading required package: MASS
Loading required package: car
Loading required package: carData

Attaching package: 'car'

The following object is masked from 'package:boot':

    logit

Loading required package: sandwich
Distribution of Main Response :  
                         "gamma" 
[1] "Estimates from the model(mu)"
y ~ crack0 + (1 | specimen)
[1] "log"
            Estimate Std. Error t-value
(Intercept)   -5.688     0.4585 -12.405
crack0         2.397     0.3852   6.222
[1] "Estimates for logarithm of lambda=var(u_mu)"
[1] "inverse-gamma"
         Estimate Std. Error t-value
specimen   -3.332     0.3443  -9.677
[1] "Estimates from the model(phi)"
phi ~ cycle + (1 | specimen)
[1] "log"
            Estimate Std. Error t-value
(Intercept)   -2.879     0.1728 -16.657
cycle         -9.052     2.2202  -4.077
[1] "Estimates for logarithm of var(u_phi)"
         Estimate Std. Error t-value
specimen   -2.101      0.803  -2.617
[1] "========== Likelihood Function Values and Condition AIC =========="
                                                         [,1]
-2ML (-2 p_v(mu),v(phi) (h))          :            -1578.2292
-2RL (-2 p_beta(mu),v(mu),beta(phi),v(phi) (h)) :  -1576.3655
cAIC                           :                   -1540.1652
Scaled Deviance                :                     222.1321
df                             :                     222.1321

dhglm documentation built on May 2, 2019, 2:08 a.m.