1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
|
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/multistart.R
\name{multistart}
\alias{multistart}
\alias{multistart.saem.mmkin}
\alias{print.multistart}
\alias{best}
\alias{best.default}
\alias{which.best}
\alias{which.best.default}
\title{Perform a hierarchical model fit with multiple starting values}
\usage{
multistart(
object,
n = 50,
cores = if (Sys.info()["sysname"] == "Windows") 1 else parallel::detectCores(),
cluster = NULL,
...
)
\method{multistart}{saem.mmkin}(object, n = 50, cores = 1, cluster = NULL, ...)
\method{print}{multistart}(x, ...)
best(object, ...)
\method{best}{default}(object, ...)
which.best(object, ...)
\method{which.best}{default}(object, ...)
}
\arguments{
\item{object}{The fit object to work with}
\item{n}{How many different combinations of starting parameters should be
used?}
\item{cores}{How many fits should be run in parallel (only on posix platforms)?}
\item{cluster}{A cluster as returned by \link[parallel:makeCluster]{parallel::makeCluster} to be used
for parallel execution.}
\item{\dots}{Passed to the update function.}
\item{x}{The multistart object to print}
}
\value{
A list of \link{saem.mmkin} objects, with class attributes
'multistart.saem.mmkin' and 'multistart'.
The object with the highest likelihood
The index of the object with the highest likelihood
}
\description{
The purpose of this method is to check if a certain algorithm for fitting
nonlinear hierarchical models (also known as nonlinear mixed-effects models)
will reliably yield results that are sufficiently similar to each other, if
started with a certain range of reasonable starting parameters. It is
inspired by the article on practical identifiabiliy in the frame of nonlinear
mixed-effects models by Duchesne et al (2021).
}
\examples{
\dontrun{
library(mkin)
dmta_ds <- lapply(1:7, function(i) {
ds_i <- dimethenamid_2018$ds[[i]]$data
ds_i[ds_i$name == "DMTAP", "name"] <- "DMTA"
ds_i$time <- ds_i$time * dimethenamid_2018$f_time_norm[i]
ds_i
})
names(dmta_ds) <- sapply(dimethenamid_2018$ds, function(ds) ds$title)
dmta_ds[["Elliot"]] <- rbind(dmta_ds[["Elliot 1"]], dmta_ds[["Elliot 2"]])
dmta_ds[["Elliot 1"]] <- dmta_ds[["Elliot 2"]] <- NULL
f_mmkin <- mmkin("DFOP", dmta_ds, error_model = "tc", cores = 7, quiet = TRUE)
f_saem_full <- saem(f_mmkin)
f_saem_full_multi <- multistart(f_saem_full, n = 16, cores = 16)
parplot(f_saem_full_multi, lpos = "bottomleft")
illparms(f_saem_full)
f_saem_reduced <- update(f_saem_full, no_random_effect = "log_k2")
illparms(f_saem_reduced)
# On Windows, we need to create a cluster first. When working with
# such a cluster, we need to export the mmkin object to the cluster
# nodes, as it is referred to when updating the saem object on the nodes.
library(parallel)
cl <- makePSOCKcluster(12)
clusterExport(cl, "f_mmkin")
f_saem_reduced_multi <- multistart(f_saem_reduced, n = 16, cluster = cl)
parplot(f_saem_reduced_multi, lpos = "topright")
}
}
\references{
Duchesne R, Guillemin A, Gandrillon O, Crauste F. Practical
identifiability in the frame of nonlinear mixed effects models: the example
of the in vitro erythropoiesis. BMC Bioinformatics. 2021 Oct 4;22(1):478.
doi: 10.1186/s12859-021-04373-4.
}
\seealso{
\link{parplot}, \link{llhist}
}
|
