R help - Jun 2008 - Coda not providing summary on mcmc object

The object is a mcmc sample from lmer. I am using R v2.7.1. Please let me
know what additional information I can provide, hopefully I am just making a
simple mistake. Thanks in advance!
> data(ratdrink, package = 'faraway')
> rd.er <- lmer(wt ~ weeks*treat + (1 | subject), data = ratdrink)
> rd.mc <- mcmcsamp(rd.er, 10000)
> library(coda)
Loading required package: lattice
> summary(rd.mc)
 Length   Class    Mode 
      1 merMCMC      S4 
> HPDinterval(rd.mc)
Error in UseMethod("HPDinterval") : 
  no applicable method for "HPDinterval"
> str(rd.mc)
Formal class 'merMCMC' [package "lme4"] with 9 slots
  ..@ Gp      : int [1:2] 0 27
  ..@ ST      : num [1, 1:10000] 1.179 0.878 0.864 0.760 0.614 ...
  ..@ call    : language lmer(formula = wt ~ weeks * treat + (1 | subject),
data = ratdrink)
  ..@ deviance: num [1:10000] 964 964 966 966 969 ...
  ..@ dims    : Named int [1:14] 1 135 6 27 1 1 1 2 5 1 ...
  .. ..- attr(*, "names")= chr [1:14] "nf" "n"
"p" "q" ...
  ..@ fixef   : num [1:6, 1:10000] 52.880 26.480  4.780 -0.794 -9.370 ...
  .. ..- attr(*, "dimnames")=List of 2
  .. .. ..$ : chr [1:6] "(Intercept)" "weeks"
"treatthiouracil"
"treatthyroxine" ...
  .. .. ..$ : NULL
  ..@ nc      : int 1
  ..@ ranef   : num[1:27, 0 ] 
  ..@ sigma   : num [1, 1:10000] 7.16 6.68 6.60 7.14 7.96
...
> search()
 [1] ".GlobalEnv"        "package:lme4"     
"package:Matrix"
"package:coda"      "package:lattice"  
"package:datasets"
"package:grDevices"
 [8] "package:graphics"  "package:stats"    
"package:utils"
"package:methods"   "Autoloads"        
"package:base"     
>  
Cheers,
Chad
-- 
View this message in context:
http://www.nabble.com/Coda-not-providing-summary-on-mcmc-object-tp18204697p18204697.html
Sent from the R help mailing list archive at Nabble.com.

smith_cc <smith_cc <at> yahoo.com> writes:
> 
> 
> The object is a mcmc sample from lmer. I am using R v2.7.1. Please let me
> know what additional information I can provide, hopefully I am just making
a
> simple mistake. Thanks in advance!
> 
> > data(ratdrink, package = 'faraway')
> > rd.er <- lmer(wt ~ weeks*treat + (1 | subject), data = ratdrink)
> > rd.mc <- mcmcsamp(rd.er, 10000)
> > library(coda)
> Loading required package: lattice
> > summary(rd.mc)
>  Length   Class    Mode 
>       1 merMCMC      S4 
> > HPDinterval(rd.mc)
> Error in UseMethod("HPDinterval") : 
>   no applicable method for "HPDinterval"
   I don't think it should necessarily have been obvious,
but help("merMCMC-class") suggests that lme4 contains
its own methods for doing HPDinterval calcs.  Furthermore,
loading coda after lme4 gives the following warning

	The following object(s) are masked from package:lme4 :

	 HPDinterval 

If you really want HPDintervals from coda, it looks like
you can do

HPDinterval(as.mcmc(as.matrix(rd.mc)))

or if you want to use lme4's HPDinterval while
coda is loaded, do

lme4::HPDinterval(rd.mc)

  (I don't quite know why lme4 is including its own
calculations rather than just allowing conversion to
coda objects -- maybe they don't want the dependency)

  If you have further questions it might make sense
to send them to r-sig-mixed ...

  Ben Bolker