R：如何从 nlmer 输出中提取随机效应

Question

library(lme4)
startvec <- c(Asym = 200, xmid = 725, scal = 350)
(nm1 <- nlmer(circumference ~ SSlogis(age, Asym, xmid, scal) ~ Asym|Tree,
              Orange, start = startvec))
> summary(nm1)$varcor
 Groups   Name Std.Dev.
 Tree     Asym 31.646  
 Residual       7.843

我正在尝试提取 Asym 随机效应的标准差（例如 31.646）。但是我已经尝试了 summary(nm1)$varcor$Std.Dev. 和 summary(nm1)$varcor[1, 3] 但没有成功。

Answer 1

stddev是一个属性

suppressWarnings(as.vector(attr(summary(nm1)$varcor$Tree, "stddev")))
#[1] 31.64632

如果我们检查

的 str

str(summary(nm1)$varcor)
#List of 1
# $ Tree: num [1, 1] 1001
#  ..- attr(*, "dimnames")=List of 2
#  .. ..$ : chr "Asym"
#  .. ..$ : chr "Asym"
#  ..- attr(*, "stddev")= Named num 31.6
#  .. ..- attr(*, "names")= chr "Asym"
#  ..- attr(*, "correlation")= num [1, 1] 1
#  .. ..- attr(*, "dimnames")=List of 2
#  .. .. ..$ : chr "Asym"
#  .. .. ..$ : chr "Asym"
# - attr(*, "sc")= num 7.84
# - attr(*, "useSc")= logi TRUE
# - attr(*, "class")= chr "VarCorr.merMod"

它是以'Tree'为单个元素的list。所以我们通过 $Tree 或 [[1]] 或 [["Tree"]] 提取 'Tree' 然后提取属性 'stddev'

R：如何从 nlmer 输出中提取随机效应

R: how to extract random effects from nlmer output

r

lme4

mixed-models