R: Plots of the results obtained by SAEM

saemix.plot.select {saemix}

R Documentation

Plots of the results obtained by SAEM

Description

Several plots (selectable by the type argument) are currently available: convergence plot, individual plots, predictions versus observations, distribution plots, residual plots, VPC.

Usage

saemix.plot.select(
  saemixObject,
  data = FALSE,
  convergence = FALSE,
  likelihood = FALSE,
  individual.fit = FALSE,
  population.fit = FALSE,
  both.fit = FALSE,
  observations.vs.predictions = FALSE,
  residuals.scatter = FALSE,
  residuals.distribution = FALSE,
  random.effects = FALSE,
  correlations = FALSE,
  parameters.vs.covariates = FALSE,
  randeff.vs.covariates = FALSE,
  marginal.distribution = FALSE,
  vpc = FALSE,
  npde = FALSE,
  ...
)

Arguments

`saemixObject`	an object returned by the `saemix` function
`data`	if TRUE, produce a plot of the data. Defaults to FALSE
`convergence`	if TRUE, produce a convergence plot. Defaults to FALSE
`likelihood`	if TRUE, produce a plot of the estimation of the LL by importance sampling. Defaults to FALSE
`individual.fit`	if TRUE, produce individual fits with individual estimates. Defaults to FALSE
`population.fit`	if TRUE, produce individual fits with population estimates. Defaults to FALSE
`both.fit`	if TRUE, produce individual fits with both individual and population estimates. Defaults to FALSE
`observations.vs.predictions`	if TRUE, produce a plot of observations versus predictions. Defaults to FALSE
`residuals.scatter`	if TRUE, produce scatterplots of residuals versus predictor and predictions. Defaults to FALSE
`residuals.distribution`	if TRUE, produce plots of the distribution of residuals. Defaults to FALSE
`random.effects`	if TRUE, produce boxplots of the random effects. Defaults to FALSE
`correlations`	if TRUE, produce a matrix plot showing the correlation between random effects. Defaults to FALSE
`parameters.vs.covariates`	if TRUE, produce plots of the relationships between parameters and covariates, using the Empirical Bayes Estimates of individual parameters. Defaults to FALSE
`randeff.vs.covariates`	if TRUE, produce plots of the relationships between random effects and covariates, using the Empirical Bayes Estimates of individual random effects. Defaults to FALSE
`marginal.distribution`	if TRUE, produce plots of the marginal distribution of the random effects. Defaults to FALSE
`vpc`	if TRUE, produce Visual Predictive Check plots. Defaults to FALSE (we suggest to use `npdeSaemix` instead)
`npde`	if TRUE, produce plots of the npde. Defaults to FALSE (deprecated in 3.0, please use `npdeSaemix` instead)
`...`	optional arguments passed to the plots

Details

This function plots different graphs related to the algorithm (convergence plots, likelihood estimation) as well as diagnostic graphs. A description is provided in the PDF documentation.

data: A spaghetti plot of the data, displaying the observed data y as a function of the regression variable (eg time for a PK application)
convergence: For each parameter in the model, this plot shows the evolution of the parameter estimate versus the iteration number
likelihood: Estimation of the likelihood estimated by importance sampling, as a function of the number of MCMC samples
individual.fit: Individual fits, using the individual parameters with the individual covariates
population.fit: Individual fits, using the population parameters with the individual covariates
both.fit: Individual fits, using the population parameters with the individual covariates and the individual parameters with the individual covariates
observations.vs.predictions: Plot of the predictions computed with the population parameters versus the observations (left), and plot of the predictions computed with the individual parameters versus the observations (right)
residuals.scatter: Scatterplot of standardised residuals versus the X predictor and versus predictions. These plots are shown for individual and population residuals, as well as for npde when they are available
residuals.distribution: Distribution of standardised residuals, using histograms and QQ-plot. These plots are shown for individual and population residuals, as well as for npde when they are available
random.effects: Boxplot of the random effects
correlations: Correlation between the random effects, with a smoothing spline
parameters.versus.covariates: Plots of the estimate of the individual parameters versus the covariates, using scatterplot for continuous covariates, boxplot for categorical covariates
randeff.versus.covariates: Plots of the estimate of the individual random effects versus the covariates, using scatterplot for continuous covariates, boxplot for categorical covariates
marginal.distribution: Distribution of each parameter in the model (conditional on covariates when some are included in the model)
npde: Plot of npde as in package npde (deprecated in 3.0, please use npdeSaemix instead)
vpc: Visual Predictive Check (we suggest to use npdeSaemix instead)

Value

None

Author(s)

Emmanuelle Comets emmanuelle.comets@inserm.fr, Audrey Lavenu, Marc Lavielle.

References

E Comets, A Lavenu, M Lavielle M (2017). Parameter estimation in nonlinear mixed effect models using saemix, an R implementation of the SAEM algorithm. Journal of Statistical Software, 80(3):1-41.

E Kuhn, M Lavielle (2005). Maximum likelihood estimation in nonlinear mixed effects models. Computational Statistics and Data Analysis, 49(4):1020-1038.

E Comets, A Lavenu, M Lavielle (2011). SAEMIX, an R version of the SAEM algorithm. 20th meeting of the Population Approach Group in Europe, Athens, Greece, Abstr 2173.

Examples


data(theo.saemix)

saemix.data<-saemixData(name.data=theo.saemix,header=TRUE,sep=" ",na=NA, 
  name.group=c("Id"),name.predictors=c("Dose","Time"),
  name.response=c("Concentration"),name.covariates=c("Weight","Sex"),
  units=list(x="hr",y="mg/L",covariates=c("kg","-")), name.X="Time")

model1cpt<-function(psi,id,xidep) { 
	  dose<-xidep[,1]
	  tim<-xidep[,2]  
	  ka<-psi[id,1]
	  V<-psi[id,2]
	  CL<-psi[id,3]
	  k<-CL/V
	  ypred<-dose*ka/(V*(ka-k))*(exp(-k*tim)-exp(-ka*tim))
	  return(ypred)
}

saemix.model<-saemixModel(model=model1cpt,
  description="One-compartment model with first-order absorption", 
  psi0=matrix(c(1.,20,0.5,0.1,0,-0.01),ncol=3, byrow=TRUE,
  dimnames=list(NULL, c("ka","V","CL"))),transform.par=c(1,1,1),
  covariate.model=matrix(c(0,1,0,0,0,0),ncol=3,byrow=TRUE),fixed.estim=c(1,1,1),
  covariance.model=matrix(c(1,0,0,0,1,0,0,0,1),ncol=3,byrow=TRUE),
  omega.init=matrix(c(1,0,0,0,1,0,0,0,1),ncol=3,byrow=TRUE),error.model="constant")

saemix.options<-list(seed=632545,save=FALSE,save.graphs=FALSE)

# Not run (strict time constraints for CRAN)
# saemix.fit<-saemix(saemix.model,saemix.data,saemix.options)

# saemix.plot.select(saemix.fit,data=TRUE,main="Spaghetti plot of data")

# Putting several graphs on the same plot
# par(mfrow=c(2,2))
# saemix.plot.select(saemix.fit,data=TRUE,vpc=TRUE,observations.vs.predictions=TRUE, new=FALSE)

[Package saemix version 3.3 Index]