[R-sig-ME] Extract variance components from lme object

BXC (Bendix Carstensen) bxc at steno.dk
Wed Sep 5 14:12:22 CEST 2007


I have fitted the following model to a three-way layout of meth, item,
repl:

lme( y ~ meth + item,
         random=list( item = pdIdent( ~ meth-1 ),
                      repl = ~1 ),
         weights = varIdent( form = ~1 | meth ),
         data=ox
         )

So I have a model with random meth-by-item and item-by-repl effects and
seprate 
residual variances for each of the two levels of meth.

How do I extract these 4 variance components to an R-structure that I
can use for 
further calculations?

The output is below.

Best
Bendix
______________________________________________

Bendix Carstensen
Senior Statistician

Steno Diabetes Center
Niels Steensens Vej 2-4
DK-2820 Gentofte
Denmark
+45 44 43 87 38 (direct)
+45 30 75 87 38 (mobile)
+45 44 43 73 13 (fax)
bxc at steno.dk   http://www.biostat.ku.dk/~bxc


Linear mixed-effects model fit by REML
  Data: ox 
  Log-restricted-likelihood: -911.7401
  Fixed: y ~ meth + item 
(Intercept)   methpulse       item2       item3       item4       item5
item6       item7       item8 
 76.0428384  -2.4704462  -7.0216227   5.1497034 -10.7281860  -1.1137199
3.1649924   9.7065633   3.5568599 
      item9      item10      item11      item12      item13      item14
item15      item16      item17 
 -4.1821374 -14.4222445  12.7503731 -47.3135668   3.3219575  -1.1293724
6.2565251  -0.5367298  13.9153464 
     item18      item19      item20      item21      item22      item23
item24      item25      item26 
  1.5322522  -2.0861271  -1.0351969   6.4653272  -0.4416475   4.5820322
8.2772197   2.1049894   2.7779659 
     item27      item28      item29      item30      item31      item32
item33      item34      item35 
-10.3186089 -10.8197187   0.7833716   2.6444795 -29.2466418   5.6528703
6.8769614   8.7365767   0.9285974 
     item36      item37      item38      item39      item40      item41
item42      item43      item44 
  3.0492155   3.6735649   7.5298316   2.7392939  -8.6159587  -0.1044011
-4.3450727 -20.7468236 -16.2943647 
     item45      item46      item47      item48      item49      item50
item51      item52      item53 
  2.0329985   4.5130501   3.4254305  -3.0309414  10.4662553 -24.8350417
-20.8508611  -0.3525354  -3.6222924 
     item54      item55      item56      item57      item58      item59
item60      item61 
  1.4299082  12.8385572   9.7971680  13.3501148  13.4953406  15.6657386
7.3963452  -1.7503731 

Random effects:
 Formula: ~meth - 1 | item
 Structure: Multiple of an Identity
          methCO methpulse
StdDev: 2.928042  2.928042

 Formula: ~1 | repl %in% item
        (Intercept) Residual
StdDev:    3.415692 2.224868

Variance function:
 Structure: Different standard deviations per stratum
 Formula: ~1 | meth 
 Parameter estimates:
      CO    pulse 
1.000000 1.795365 
Number of Observations: 354
Number of Groups: 
          item repl %in% item 
            61            177 
>




More information about the R-sig-mixed-models mailing list