Overview

Description

buildAll builds the complete statistical model by iteratively calling functions buildmlx and buildVar.

Usage

buildAll <- function(project=NULL, final.project=NULL, model="all", prior=NULL, weight=NULL, coef.w1=0.5, cv.min=0.001, fError.min=1e-3,
                     paramToUse="all", covToTest="all", covToTransform="none", center.covariate=FALSE, 
                     criterion="BICc", linearization=FALSE, ll=T, test=T, direction=NULL, steps=1000,
                     max.iter=20, explor.iter=2, seq.cov=FALSE, seq.corr=TRUE, seq.cov.iter=0, 
                     p.max=0.1, p.min=c(0.075, 0.05, 0.1), print=TRUE, nb.model=1,
                     fix.param1=NULL, fix.param0=NULL, remove=T, add=T, delta=c(30,10,5), 
                     omega.set=NULL, pop.set1=NULL, pop.set2=NULL)

Arguments

project

a string: the initial Monolix project

final.project

the final Monolix project (default is the original project)a string: the final Monolix project (default adds “_buildAll” to the original project)

model

the components of the model to optimize c(“residualError”, “covariate”, “correlation”), (default=“all”)

prior

list of prior probabilities for each component of the model (default=NULL)

weight

list of penalty weights for each component of the model (default=NULL)

coef.w1

multiplicative weight coefficient used for the first iteration only (default=0.5)

cv.min

value of the coefficient of variation below which an individual parameter is considered fixed (default=0.001)

fError.min

minimum fraction of residual variance for combined error model (default = 1e-3)

paramToUse

list of parameters possibly function of covariates (default=“all”)

covToTest

components of the covariate model that can be modified (default=“all”)

covToTransform

list of (continuous) covariates to be log-transformed (default=“none”)

center.covariate

TRUE/{FALSE} center the covariates of the final model (default=FALSE)

criterion

penalization criterion to optimize c(“AIC”, “BIC”, {“BICc”}, gamma)

linearization

TRUE/{FALSE} whether the computation of the likelihood is based on a linearization of the model (default=FALSE)

{TRUE}/FALSE compute the observe likelihood and the criterion to optimize at each iteration (default=TRUE)

test

{TRUE}/FALSE perform additional statistical tests for building the model (default=TRUE)

direction

method for covariate search c({“full”}, “both”, “backward”, “forward”), (default=“full” or “both”)

steps

maximum number of iteration for stepAIC (default=1000)

max.iter

maximum number of SAMBA iterations (default=20)

exp.iter

number of iterations during the exploratory phase of SAMBA (default=2)

seq.cov

TRUE/{FALSE} whether the covariate model is built before the correlation model (default=FALSE)

seq.corr

{TRUE}/FALSE whether the correlation model is built iteratively (default=TRUE)

seq.cov.iter

number of iterations before building the correlation model (only when seq.cov=F, default=0)

p.max

maximum p-value used for removing non significant relationships between covariates and individual parameters (default=0.1)

p.min

vector of 3 minimum p-values used for testing the components of a new model (default=c(0.075, 0.05, 0.1))

{TRUE}/FALSE display the results (default=TRUE)

nb.model

number of models to display at each iteration (default=1)

fix.param1

parameters with variability that cannot be removed (default=NULL)

fix.param0

parameters without variability that cannot be added (default=NULL)

remove

{TRUE}/FALSE try to remove random effects (default=TRUE)

add

{TRUE}/FALSE try to add random effects (default=TRUE)

delta

maximum difference in criteria for testing a new model (default=c(30,10,5))

omega.set

settings to define how a variance varies during iterations of SAEM

pop.set1

Monolix settings 1

pop.set2

Monolix settings 2

Example

Using version <= 4.0.2 of Rsmlx requires to load explicitly lixoftConnectors. This package is automatically loaded with Rsmlx 4.0.3.

library(lixoftConnectors)
initializeLixoftConnectors(software="monolix")
library(Rsmlx)

We select a Monolix project

project <- "projects/simulatedPK2.mlxtran"

Example using the defaults settings of buildmlx and buildVar:

buildAll.res1 <- buildAll(project)

Selected model:

loadProject(buildAll.res1$project)
getIndividualParameterModel()

## $name
## [1] "ka" "Cl" "V1" "Q"  "V2"
## 
## $distribution
##          ka          Cl          V1           Q          V2 
## "logNormal" "logNormal" "logNormal" "logNormal" "logNormal" 
## 
## $limits
## named list()
## 
## $formula
## [1] "log(ka) = log(ka_pop) + eta_ka\nlog(Cl) = log(Cl_pop) + beta_Cl_X03*X03 + eta_Cl\nlog(V1) = log(V1_pop) + beta_V1_X01*X01 + beta_V1_X02*X02 + eta_V1\nlog(Q) = log(Q_pop) + beta_Q_X01*X01\nlog(V2) = log(V2_pop)\nCorrelations\n\tID : {Cl, V1}\n"
## 
## $variability
## $variability$id
##    ka    Cl    V1     Q    V2 
##  TRUE  TRUE  TRUE FALSE FALSE 
## 
## 
## $covariateModel
## $covariateModel$ka
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$Cl
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$V1
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
##  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$Q
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
##  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$V2
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## 
## $correlationBlocks
## $correlationBlocks$id
## $correlationBlocks$id[[1]]
## [1] "Cl" "V1"

Example using several settings of buildmlx and buildVar:

buildAll.res2 <- buildAll(project, final.project="projects/buildAll2.mlxtran", 
                          paramToUse=c("Cl", "V1", "Q"), covToTest=c("X01", "X02", "X03", "X04"),
                          fix.param0="Q", fix.param1=c("ka", "Cl"))

Selected model:

loadProject(buildAll.res2$project)
getIndividualParameterModel()

## $name
## [1] "ka" "Cl" "V1" "Q"  "V2"
## 
## $distribution
##          ka          Cl          V1           Q          V2 
## "logNormal" "logNormal" "logNormal" "logNormal" "logNormal" 
## 
## $limits
## named list()
## 
## $formula
## [1] "log(ka) = log(ka_pop) + eta_ka\nlog(Cl) = log(Cl_pop) + beta_Cl_X03*X03 + eta_Cl\nlog(V1) = log(V1_pop) + beta_V1_X01*X01 + beta_V1_X02*X02 + eta_V1\nlog(Q) = log(Q_pop) + beta_Q_X01*X01\nlog(V2) = log(V2_pop)\nCorrelations\n\tID : {Cl, V1}\n"
## 
## $variability
## $variability$id
##    ka    Cl    V1     Q    V2 
##  TRUE  TRUE  TRUE FALSE FALSE 
## 
## 
## $covariateModel
## $covariateModel$ka
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$Cl
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$V1
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
##  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$Q
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
##  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$V2
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## 
## $correlationBlocks
## $correlationBlocks$id
## $correlationBlocks$id[[1]]
## [1] "Cl" "V1"

Following what is done with buildmlxand buildVar, prior information about the statistical model can be introduced either by defining prior probabilities or introducing a weighting for the penalty term.

In this example, relationships between covariates and parameters are strongly penalized (weight=5) while correlations are lightly penalized (weight=0.1).

Different weights are used for the variances: variability of Q is privileged while that of V1 is strongly penalized.

buildAll.res3 <- buildAll(project, weight=list(covariate=5, correlation=0.3, variance=c(Q=0.05, V1=20)))

print(buildAll.res3$variability.model)

## NULL

loadProject(buildAll.res3$project)
getIndividualParameterModel()

## $name
## [1] "ka" "Cl" "V1" "Q"  "V2"
## 
## $distribution
##          ka          Cl          V1           Q          V2 
## "logNormal" "logNormal" "logNormal" "logNormal" "logNormal" 
## 
## $limits
## named list()
## 
## $formula
## [1] "log(ka) = log(ka_pop) + eta_ka\nlog(Cl) = log(Cl_pop) + beta_Cl_X03*X03 + eta_Cl\nlog(V1) = log(V1_pop) + beta_V1_X01*X01 + beta_V1_X02*X02 + eta_V1\nlog(Q) = log(Q_pop) + beta_Q_X01*X01\nlog(V2) = log(V2_pop)\nCorrelations\n\tID : {Cl, V1}\n"
## 
## $variability
## $variability$id
##    ka    Cl    V1     Q    V2 
##  TRUE  TRUE  TRUE FALSE FALSE 
## 
## 
## $covariateModel
## $covariateModel$ka
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$Cl
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$V1
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
##  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$Q
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
##  TRUE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## $covariateModel$V2
##   X01   X02   X03   X04   X05   X06   X07   X08   X09   X10 
## FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE 
## 
## 
## $correlationBlocks
## $correlationBlocks$id
## $correlationBlocks$id[[1]]
## [1] "Cl" "V1"

Automatic complete statistical model building

Marc Lavielle
June 14th, 2022

Overview

Description

Usage

Arguments

Example

Automatic complete statistical model building

Marc Lavielle June 14th, 2022

Overview

Description

Usage

Arguments

Example

Marc Lavielle
June 14th, 2022