Creating an IQRnlmeProject object

This function takes and IQRnlmeEst object and other optional input arguments and exports an NMLE project to a folder. It is this folder that we call an "IQRnlmeProject" object and it is tool specific (e.g. NONMEM, MONOLIX, or NLMIXR). The original IQRmodel used here will be exported to the project folder as well as "model.txt" The project folder will contain in addition the IQRnlmeEst object as a project.est file.

IQRnlmeProject(
  est,
  projectPath,
  comment = "",
  tool = "MONOLIX",
  toolVersion = NULL,
  multiTestN = 1,
  multiTestSD = 0.5,
  FLAGanalytic = TRUE,
  keepProjectFolder = FALSE,
  algOpt.SEED = 123456,
  algOpt.K1 = 500,
  algOpt.K2 = 200,
  algOpt.NRCHAINS = NA,
  algOpt.NONMEM.METHOD = "SAEM",
  algOpt.NONMEM.MAXEVAL = 9999,
  algOpt.NONMEM.SIGDIGITS = 3,
  algOpt.NONMEM.PRINT = 1,
  algOpt.NONMEM.COVSTEP_MATRIX = "S",
  algOpt.NONMEM.ADVAN7 = TRUE,
  algOpt.NONMEM.N1 = 1000,
  algOpt.NONMEM.TOL = 6,
  algOpt.NONMEM.SIGL = NULL,
  algOpt.NONMEM.M4 = FALSE,
  algOpt.NONMEM.FOCEIOFV = FALSE,
  algOpt.NONMEM.IMPORTANCESAMPLING = TRUE,
  algOpt.NONMEM.IMP_ITERATIONS = 10,
  algOpt.NONMEM.ITS = TRUE,
  algOpt.NONMEM.ITS_ITERATIONS = 10,
  algOpt.NONMEM.WRES = NULL,
  algOpt.NONMEM.PRED = NULL,
  algOpt.NONMEM.RES = NULL,
  algOpt.MONOLIX.individualParameters = "conditionalMode",
  algOpt.MONOLIX.indivMCMClength = 50,
  algOpt.MONOLIX.indivNsim = 10,
  algOpt.MONOLIX.indivRatio = 0.05,
  algOpt.MONOLIX.logLikelihood = "Linearization",
  algOpt.MONOLIX.fim = "Linearization",
  algOpt.MONOLIX.variability = "FirstStage",
  algOpt.MONOLIX.startTime = NULL,
  algOpt.MONOLIX.STIFF = TRUE,
  algOpt.NLMIXR.method = "SAEM",
  algOpt.NLMIXR.control = NULL,
  verbose = TRUE
)

Arguments

est

IQRnlmeEst object
projectPath

Path where to generate the NLME project
comment

Character string. Allows the user to enter some information about the model. This can be displayed in summary tables. It is stored in the model in the metadata section
tool

character string with tool name ("NONMEM", "MONOLIX", or "NLMIXR")
toolVersion

character string with the tool version name as defined in the setup_options_IQRtools.R file (can be set with setup_IQRtools() function). By default the first option defined in the respective lists in setup_options_IQRtools.R is used.
multiTestN

Doing robustness analysis - number of models to generate with different initial guesses for fixed effects (randomly generated based on POPvalues0). Only fixed effects which are considered for estimation will be randomized. If multiTestN=1 (default) then no randomization is done.
multiTestSD

Standard deviation to use to add noise__ to the initial parameter guesses (default=0.5 (50%CV)) Normal: Parameter_guess + multiTestSDParameter_guessrandomNumbers(0-1) LogNormal: Parameter_guess * exp(multiTestSD*randomNumbers(0-1)) LogitNormal: Similar and between 0-1
FLAGanalytic

TRUE: use analytic solution if possible, FALSE: always use ODE based solution.
keepProjectFolder

FALSE: erases projectFolder before creating new. TRUE: keeps projectFolder contents when creating new NLME project files.
algOpt.SEED

Used for NONMEM SAEM, MONOLIX, and NLMIXR SAEM
algOpt.K1

First iterations for SAEM. Same for NONMEM SAEM, MONOLIX, and NLMIXR SAEM
algOpt.K2

Second iterations for SAEM. Same for NONMEM SAEM, MONOLIX, and NLMIXR SAEM
algOpt.NRCHAINS

Used for NONMEM SAEM, MONOLIX, and NLMIXR SAEM. In case of few subjects in the dataset and use of the SAEM method it is important to adjust this parameter. It should be chosen such that the NRCHAINS * number of subjects in the dataset is >50. The user can set this value but by default it is chosen automatically such that NRCHAINS * number of subjects >= 50.
algOpt.NONMEM.METHOD

'FO','FOCE','FOCEI','SAEM','BAYES' (ignored when using MONOLIX)
algOpt.NONMEM.MAXEVAL

NONMEM option
algOpt.NONMEM.SIGDIGITS

NONMEM option
algOpt.NONMEM.PRINT

NONMEM option
algOpt.NONMEM.COVSTEP_MATRIX

Option to set tzhe MATRIX type for the $COV step in NONMEM. Allowed are "S", "R", and "RSR". "RSR" translates to unspecified MATRIX. IQRtools uses S as default. It works (for larger number of subjects - and for small number MONOLIX is anyway better)
algOpt.NONMEM.ADVAN7

TRUE: use ADVAN7 in case that model can be solved analytically. FALSE: use ADVAN5. This is only used in case of NONMEM. ADVAN5 handles complex eigenvalues. ADVAN7 real ones and can be faster.
algOpt.NONMEM.N1

ESAMPLE setting for tables.
algOpt.NONMEM.TOL

Tolerance settings for ADVAN13. For 1e-6 set TOL=6
algOpt.NONMEM.SIGL

Number of significant digits for LL calculation. Default: undefined If SIGL defined, Only used in gradient based methods. Good to set SIGDIG to <=SIGL/3.
algOpt.NONMEM.M4

if CENS=1 entries present in data and M4=TRUE then the M4 method is used, otherwise the M3 method. this requires DV to be set to LLOQ and CENS=1 if the observation is BLLOQ.
algOpt.NONMEM.FOCEIOFV

Determine the objective function value using an FOCE step with MAXEVAL=0. Experts agree that the IMP method with MAPITER=0 leads to better objective functions. However, in practice this method suffers from non reproducibility when running on multiple cores. Even at identical population parameter estimates the obtained OFVs can differ greatly. Therefor, instead of using the IMP method after an SAEM estimation, this option allows to perform a FOCEI step with MAXEVAL=0. This option cannot be used in combination with algOpt.NONMEM.IMPORTANCESAMPLING.
algOpt.NONMEM.IMPORTANCESAMPLING

Determine the objective function value with importance sampling
algOpt.NONMEM.IMP_ITERATIONS

Number of iterations for importance sampling
algOpt.NONMEM.ITS

Allow to run an ITS method as first method befor all other methods (METHOD) ITS = FALSE or TRUE. ITS=TRUE only accepted if not FO!
algOpt.NONMEM.ITS_ITERATIONS

Number of ITS iterations (NONMEM only)
algOpt.NONMEM.WRES

String with name of weighted residuals to use in diagnostics (by default the standard ones are used based the estimation method and on the NONMEM manual). Example: "CWRES", "CWRESI", "EWRES", etc.
algOpt.NONMEM.PRED

String with name of predictions to use in diagnostics (by default the standard ones are used based the estimation method and on the NONMEM manual). Example: "CPREDI", "EPRED", "NPRED", etc.
algOpt.NONMEM.RES

String with name of residuals to use in diagnostics (by default the standard ones are used based the estimation method and on the NONMEM manual). Example: "CRESI", "ERES", "NPRES", etc.
algOpt.MONOLIX.individualParameters

MONOLIX option. 'conditionalMode' (default), 'conditionalMean', or 'All'
algOpt.MONOLIX.indivMCMClength

MONOLIX option if individualParameters 'conditionalMean' or 'All'. Interval length of the MCMC convergence assessment for the conditional distribution
algOpt.MONOLIX.indivNsim

MONOLIX option if individualParameters 'conditionalMean' or 'All'. Number of simulated individual parameters per individual
algOpt.MONOLIX.indivRatio

MONOLIX option if individualParameters 'conditionalMean' or 'All'. Relative interval width (between 0 and 1)
algOpt.MONOLIX.logLikelihood

MONOLIX option. 'Linearization' (default) or 'ImportanceSampling'
algOpt.MONOLIX.fim

MONOLIX option. 'Linearization' (default) or 'StochasticApproximation'
algOpt.MONOLIX.variability

MONOLIX option. 'no', 'Decreasing', 'FirstStage' (default), 'Mixed'
algOpt.MONOLIX.startTime

MONOLIX option. Start time of integration. default: NULL (not set).
algOpt.MONOLIX.STIFF

MONOLIX option. Use stiff solver (TRUE-default) or non-stiff (FALSE).
algOpt.NLMIXR.method

NLMIXR method ("SAEM" or "NLME" or "FOCEI")
algOpt.NLMIXR.control

Control settings for chosen method. Use "nlmixr::saemControl()" or "nlmixr::nlmeControl()" or "nlmixr::foceiControl()", depending on chosen method. If SAEM is chosen and nlmixr::saemControl(...) is defined, the options algOpt.SEED, algOpt.K1, algOpt.K2, and algOpt.NRCHAINS will be overwritten if the respective arguments (seed, nBurn, nEm, nmc) are defined in the nlmixr::saemControl(...) call.
verbose

Logical. If FALSE then text output in console is avoided (except warnings and errors).

Value

An IQRnlmeProject object

See also

Other IQRnlmeProject: IQRnlmeEst(), addCovariateToModelSpec_IQRest(), addPar_modelSpec(), as_IQRnlmeProjectMulti(), as_IQRnlmeProject(), bootstrap_IQRnlmeProject(), compareModels_IQRnlmeProjectMulti(), convertETAINDIVPRED_IQRnlmeProject(), covariateEffect_IQRnlmeProject(), createDataVPC_IQRnlmeProject(), data_IQRest(), dosing_IQRest(), duplicate_IQRnlmeProject(), eigenvalues_IQRnlmeProject(), exportVirtualSubjects_IQRnlmeProject(), getData_IQRnlmeProject(), getETAs_IQRnlmeProject(), getEst_IQRnlmeProject(), getHeader_IQRnlmeProject(), getIndivParameters_IQRnlmeProject(), getIndivPredictions_IQRnlmeProject(), getModel_IQRnlmeProject(), getObsPred_IQRnlmeProject(), getPopParameters_IQRnlmeProject(), getResults_IQRnlmeProjectMulti(), getResults_IQRnlmeProject(), hasrun_IQRnlmeProjectMulti(), hasrun_IQRnlmeProject(), informationContent_IQRnlmeProject(), is_IQRnlmeEst(), is_IQRnlmeProjectMulti(), is_IQRnlmeProject(), is_MONOLIX_IQRnlmeProject(), is_NLMIXR_IQRnlmeProject(), is_NONMEM_IQRnlmeProject(), modelSpec_IQRest(), outlier_IQRnlmeProject(), plot.IQRnlmeProjectMulti(), plot.IQRnlmeProject(), plotConvergence_IQRnlmeProject(), plotETACOV_IQRnlmeProject(), plotETA_IQRnlmeProject(), plotGOF_IQRnlmeProject(), plotINDIVSIM_IQRnlmeProject(), plotINDIV_IQRnlmeProject(), plotVPC_IQRdataVPC(), pred_IQRnlmeProject(), print.IQRnlmeEst(), print.IQRnlmeProjectMulti(), print.IQRnlmeProject(), print_modelSpec(), run_IQRnlmeProjectMulti(), run_IQRnlmeProject(), sample_IQRnlmeProject(), summary.IQRnlmeProjectMulti(), summary.IQRnlmeProject(), summaryComments_IQRnlmeProjectMulti(), summaryCorrelations_IQRnlmeProjectMulti(), summaryCovariates_IQRnlmeProjectMulti(), summaryParameters_IQRnlmeProjectMulti(), vpc_IQRnlmeProject()