GitHub - nlmixr2/nlmixr2save: nlmixr2save is a package to save a nlmixr2 fit in a format that is (mostly) readable by any tool.

# nlmixr2save

The goal of nlmixr2save is to save a nlmixr2 fit in file formats that can be read outside R, avoiding .rds or other formats as much as possible.

Installation

You can install the development version of nlmixr2save from GitHub with:

# install.packages("remotes")
remotes::install_github("nlmixr2/nlmixr2save")

Example

This is a basic example which shows you how to solve a common problem:

library(nlmixr2)
#> ── Attaching packages ───────────────────────────────────────── nlmixr2 5.0.0 ──
#> ✔ lotri         1.0.5          ✔ monolix2rx    0.0.6     
#> ✔ nlmixr2data   2.0.9          ✔ nlmixr2lib    0.3.2     
#> ✔ nlmixr2est    5.0.2.9000     ✔ nlmixr2rpt    0.2.2     
#> ✔ nlmixr2extra  5.0.0          ✔ nonmem2rx     0.1.9     
#> ✔ nlmixr2plot   5.0.1          ✔ posologyr     1.2.8     
#> ✔ rxode2        5.1.0          ✔ shinyMixR     0.5.3     
#> ✔ babelmixr2    0.1.11         ✔ xpose.nlmixr2 0.4.2     
#> ✔ ggPMX         1.3.2
#> ── Optional Packages Loaded/Ignored ─────────────────────────── nlmixr2 5.0.0 ──
#> ✔ babelmixr2        ✔ nonmem2rx    
#> ✔ ggPMX             ✔ posologyr    
#> ✔ monolix2rx        ✔ shinyMixR    
#> ✔ nlmixr2lib        ✔ xpose.nlmixr2
#> ✔ nlmixr2rpt
#> ── Conflicts ───────────────────────────────────────────── nlmixr2conflicts() ──
#> ✖ rxode2::boxCox()     masks nlmixr2est::boxCox()
#> ✖ rxode2::yeoJohnson() masks nlmixr2est::yeoJohnson()
library(nlmixr2save) # eventually nlmixr2 will include this if available
one.cmt <- function() {
  ini({
    ## You may label each parameter with a comment
    tka <- 0.45 # Log Ka
    tcl <- log(c(0, 2.7, 100)) # Log Cl
    ## This works with interactive models
    ## You may also label the preceding line with label("label text")
    tv <- 3.45; label("log V")
    ## the label("Label name") works with all models
    eta.ka ~ 0.6
    eta.cl ~ 0.3
    eta.v ~ 0.1
    add.sd <- 0.7
  })
  model({
    ka <- exp(tka + eta.ka)
    cl <- exp(tcl + eta.cl)
    v <- exp(tv + eta.v)
    linCmt() ~ add(add.sd)
  })
}

fitF <- suppressMessages(nlmixr(one.cmt, theo_sd, est="focei",
                                control=list(print=0, compress=FALSE)))
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> 
#> calculating covariance matrix
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00 
#> done

saveFit(fitF)
#> ℹ saving fit item: .rownum
#> ℹ saving fit item: AIC
#> ℹ saving fit item: aqHi
#> ℹ saving fit item: aqLow
#> ℹ saving fit item: aqn
#> ℹ saving fit item: BIC
#> ℹ saving fit item: censInformation
#> ℹ saving fit item: cholR
#> ℹ saving fit item: cholS
#> ℹ saving fit item: conditionNumberCor
#> ℹ saving fit item: conditionNumberCov
#> ℹ saving fit item: convergence
#> ℹ saving fit item: cov
#> ℹ saving fit item: covLvl
#> ℹ saving fit item: covMethod
#> ℹ saving fit item: covR
#> ℹ saving fit item: covRS
#> ℹ saving fit item: covS
#> ℹ saving fit item: eigenCor
#> ℹ saving fit item: eigenCov
#> ℹ saving fit item: eigenVecCor
#> ℹ saving fit item: eigenVecCov
#> ℹ saving fit item: est
#> ℹ saving fit item: etaObf
#> ℹ saving fit item: extra
#> ℹ saving fit item: fixef
#> ℹ saving fit item: foceiControl0
#> ℹ saving fit item: foceiModel
#> ℹ saving fit item: fullCor
#> ℹ saving fit item: iniDf0
#> ℹ saving fit item: llikObs
#> ℹ saving fit item: logLik
#> ℹ saving fit item: message
#> ℹ saving fit item: method
#> ℹ saving fit item: mixIdx
#> ℹ saving fit item: nAGQ
#> ℹ saving fit item: nEstOmega
#> ℹ saving fit item: nobs
#> ℹ saving fit item: nsub
#> ℹ saving fit item: objDf
#> ℹ saving fit item: objective
#> ℹ saving fit item: ofvType
#> ℹ saving fit item: omega
#> ℹ saving fit item: optReturn
#> ℹ saving fit item: origData
#> ℹ saving fit item: parFixed
#> ℹ saving fit item: parFixedDf
#> ℹ saving fit item: parHistData
#> ℹ saving fit item: phiC
#> ℹ saving fit item: phiH
#> ℹ saving fit item: qfirst
#> ℹ saving fit item: qw
#> ℹ saving fit item: qx
#> ℹ saving fit item: R
#> ℹ saving fit item: R.0
#> ℹ saving fit item: R.E
#> ℹ saving fit item: R.pd
#> ℹ saving fit item: ranef
#> ℹ saving fit item: Rinv
#> ℹ saving fit item: runInfo
#> ℹ saving fit item: S
#> ℹ saving fit item: S.E
#> ℹ saving fit item: S.pd
#> ℹ saving fit item: S0
#> ℹ saving fit item: scaleInfo
#> ℹ saving fit item: sessioninfo
#> ℹ saving fit item: shrink
#> ℹ saving fit item: Sper
#> ℹ saving fit item: table
#> ℹ saving fit item: time
#> ℹ saving fit item: tolFactor
#> ℹ saving fit item: ui
#> ℹ zipping fit files
#> ℹ removing unzipped fit files

# You can also load this with `loadFit(fitF)`

print(loadFit(fitF))
#> ℹ loading fit from fitF.R
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ removing unzipped fit files
#> ── nlmixr² FOCEi (outer: nlminb) ──
#> 
#>           OBJF      AIC      BIC Log-likelihood Condition#(Cov) Condition#(Cor)
#> FOCEi 116.8044 373.4041 393.5838      -179.7021        68.53386        9.375135
#> 
#> ── Time (sec $time): ──
#> 
#>            setup optimize covariance table compress    other
#> elapsed 0.003785 0.441136   0.441138 0.072    0.001 5.195941
#> 
#> ── Population Parameters ($parFixed or $parFixedDf): ──
#> 
#>        Parameter  Est.     SE %RSE Back-transformed(95%CI) BSV(CV%) Shrink(SD)%
#> tka              0.463  0.195 42.2       1.59 (1.08, 2.33)     70.7      2.03% 
#> tcl               1.01 0.0751 7.42       2.75 (2.38, 3.19)     26.7      3.89% 
#> tv         log V  3.46 0.0436 1.26       31.8 (29.2, 34.7)     14.0      10.5% 
#> add.sd           0.694                               0.694                     
#>  
#>   Covariance Type ($covMethod): r,s
#>   Some strong fixed parameter correlations exist ($cor) :
#>     cor:tcl,tka  cor:tv,tka  cor:tv,tcl 
#>      0.158       0.421       0.744  
#>  
#> 
#>   No correlations in between subject variability (BSV) matrix
#>   Full BSV covariance ($omega) or correlation ($omegaR; diagonals=SDs) 
#>   Distribution stats (mean/skewness/kurtosis/p-value) available in $shrink 
#>   Information about run found ($runInfo):
#>    • gradient problems with initial estimate and covariance; see $scaleInfo 
#>    • ETAs were reset to zero during optimization; (Can control by foceiControl(resetEtaP=.)) 
#>    • initial ETAs were nudged; (can control by foceiControl(etaNudge=., etaNudge2=)) 
#>   Censoring ($censInformation): No censoring
#>   Minimization message ($message):  
#>     relative convergence (4) 
#> 
#> ── Fit Data (object is a modified tibble): ──
#> # A tibble: 132 × 22
#>      ID  TIME    DV  PRED    RES   WRES IPRED   IRES  IWRES CPRED   CRES  CWRES
#>   <int> <dbl> <dbl> <dbl>  <dbl>  <dbl> <dbl>  <dbl>  <dbl> <dbl>  <dbl>  <dbl>
#> 1     1  0     0.74  0     0.74   1.07   0     0.74   1.07   0     0.74   1.07 
#> 2     1  0.25  2.84  3.26 -0.419 -0.223  3.85 -1.01  -1.45   3.21 -0.374 -0.175
#> 3     1  0.57  6.57  5.82  0.746  0.299  6.78 -0.215 -0.310  5.77  0.802  0.289
#> # ℹ 129 more rows
#> # ℹ 10 more variables: eta.ka <dbl>, eta.cl <dbl>, eta.v <dbl>, depot <dbl>,
#> #   central <dbl>, ka <dbl>, cl <dbl>, v <dbl>, tad <dbl>, dosenum <int>

This creates a “.zip” file that includes a “fitF.R” file. This fit builds the fit from (mostly) csv and R source files.

The reason for this extra package saving models:

The model properties and results can be examined outside of R
This model can be saved without worrying how data are serialized and/or compressed
This saved model version should be independent of the version of rxode2 and nlmixr2, whereas other methods of saving are dependent on the version of nlmixr2/rxode2

In the future, this could be used in conjunction with vroom, altrep and other similar tools to partially load fit without loading the corresponding datasets.

In addition to introducing the R-independent zip file for nlmixr2 fits, this also introduced the cache assignment operator := which will cache functions calls when assigning them.

In the case of nlmixr2 fits, these cached values will be the zip files described above; Otherwise, this will be a rds file that has the object and the md5 of the arguments in a list.

If the object is NOT a function, it cannot load the cache, but may have some side effects.

In the case of a nlmixr2 fit, you could assign it as follows

fit2 := fit # The fit2.zip is created to save the fit

Here are some annotated examples of the cache assignment operator in pratice:

# You can also automatically load a fit if the arguments for nlmixr2 do not change
# by using the := cached assignment operator introduced in this package

fitF := nlmixr(one.cmt, theo_sd, est="focei", control=list(print=0, compress=FALSE))
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ loading fit from fitF.R
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ removing unzipped fit files
#> ℹ fit in fitF.zip does not match current fit; removing and refitting
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> calculating covariance matrix
#> [====|====|====|====|====|====|====|====|====|====done
#> → Calculating residuals/tables
#> ✔ done
#> ℹ saving fit item: .rownum
#> ℹ saving fit item: AIC
#> ℹ saving fit item: aqHi
#> ℹ saving fit item: aqLow
#> ℹ saving fit item: aqn
#> ℹ saving fit item: BIC
#> ℹ saving fit item: censInformation
#> ℹ saving fit item: cholR
#> ℹ saving fit item: cholS
#> ℹ saving fit item: conditionNumberCor
#> ℹ saving fit item: conditionNumberCov
#> ℹ saving fit item: convergence
#> ℹ saving fit item: cov
#> ℹ saving fit item: covLvl
#> ℹ saving fit item: covMethod
#> ℹ saving fit item: covR
#> ℹ saving fit item: covRS
#> ℹ saving fit item: covS
#> ℹ saving fit item: eigenCor
#> ℹ saving fit item: eigenCov
#> ℹ saving fit item: eigenVecCor
#> ℹ saving fit item: eigenVecCov
#> ℹ saving fit item: est
#> ℹ saving fit item: etaObf
#> ℹ saving fit item: extra
#> ℹ saving fit item: fixef
#> ℹ saving fit item: foceiControl0
#> ℹ saving fit item: foceiModel
#> ℹ saving fit item: fullCor
#> ℹ saving fit item: iniDf0
#> ℹ saving fit item: llikObs
#> ℹ saving fit item: logLik
#> ℹ saving fit item: message
#> ℹ saving fit item: method
#> ℹ saving fit item: mixIdx
#> ℹ saving fit item: nAGQ
#> ℹ saving fit item: nEstOmega
#> ℹ saving fit item: nlmixr2save
#> ℹ saving fit item: nlmixr2saveOrig
#> ℹ saving fit item: nobs
#> ℹ saving fit item: nsub
#> ℹ saving fit item: objDf
#> ℹ saving fit item: objective
#> ℹ saving fit item: ofvType
#> ℹ saving fit item: omega
#> ℹ saving fit item: optReturn
#> ℹ saving fit item: origData
#> ℹ saving fit item: parFixed
#> ℹ saving fit item: parFixedDf
#> ℹ saving fit item: parHistData
#> ℹ saving fit item: phiC
#> ℹ saving fit item: phiH
#> ℹ saving fit item: qfirst
#> ℹ saving fit item: qw
#> ℹ saving fit item: qx
#> ℹ saving fit item: R
#> ℹ saving fit item: R.0
#> ℹ saving fit item: R.E
#> ℹ saving fit item: R.pd
#> ℹ saving fit item: ranef
#> ℹ saving fit item: Rinv
#> ℹ saving fit item: runInfo
#> ℹ saving fit item: S
#> ℹ saving fit item: S.E
#> ℹ saving fit item: S.pd
#> ℹ saving fit item: S0
#> ℹ saving fit item: scaleInfo
#> ℹ saving fit item: sessioninfo
#> ℹ saving fit item: shrink
#> ℹ saving fit item: Sper
#> ℹ saving fit item: table
#> ℹ saving fit item: time
#> ℹ saving fit item: tolFactor
#> ℹ saving fit item: ui
#> ℹ zipping fit files
#> ℹ removing unzipped fit files

# Now using the same arguments it loads
fitF := nlmixr(one.cmt, theo_sd, est="focei", control=list(print=0, compress=FALSE))
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ loading fit from fitF.R
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ removing unzipped fit files

# But if we use different arguments, say using saem, it will rerun.
fitF := nlmixr(one.cmt, theo_sd, est="saem", control=list(print=0, compress=FALSE))
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ loading fit from fitF.R
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> ℹ removing unzipped fit files
#> ℹ fit in fitF.zip does not match current fit; removing and refitting
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments
#> → loading into symengine environment...
#> → pruning branches (`if`/`else`) of saem model...
#> ✔ done
#> → finding duplicate expressions in saem model...
#> [====|====|====|====|====|====|====|====|====|====
#> ====|====|====|====|====|====|====|====|====|====] 0:00:00
#> ✔ done
#> ℹ calculate uninformed etas
#> ℹ done
#> Calculating covariance matrix
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00
#> → loading into symengine environment...
#> → pruning branches (`if`/`else`) of saem model...
#> ✔ done
#> → finding duplicate expressions in saem predOnly model 0...
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00
#> → finding duplicate expressions in saem predOnly model 1...
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00
#> → finding duplicate expressions in saem predOnly model 2...
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00
#> → optimizing duplicate expressions in saem predOnly model 2...
#> [====|====|====|====|====|====|====|====|====|====] 0:00:00
#> ✔ done
#> → Calculating residuals/tables
#> ✔ done
#> ℹ saving fit item: .likTime
#> ℹ saving fit item: .rownum
#> ℹ saving fit item: adjObf
#> ℹ saving fit item: AIC
#> ℹ saving fit item: aqHi
#> ℹ saving fit item: aqLow
#> ℹ saving fit item: aqn
#> ℹ saving fit item: BIC
#> ℹ saving fit item: censInformation
#> ℹ saving fit item: conditionNumberCor
#> ℹ saving fit item: conditionNumberCov
#> ℹ saving fit item: cov
#> ℹ saving fit item: covLvl
#> ℹ saving fit item: covMethod
#> ℹ saving fit item: eigenCor
#> ℹ saving fit item: eigenCov
#> ℹ saving fit item: eigenVecCor
#> ℹ saving fit item: eigenVecCov
#> ℹ saving fit item: est
#> ℹ saving fit item: etaObf
#> ℹ saving fit item: extra
#> ℹ saving fit item: fixef
#> ℹ saving fit item: foceiControl0
#> ℹ saving fit item: fullCor
#> ℹ saving fit item: iniDf0
#> ℹ saving fit item: logLik
#> ℹ saving fit item: message
#> ℹ saving fit item: method
#> ℹ saving fit item: mixIdx
#> ℹ saving fit item: nAGQ
#> ℹ saving fit item: nEstOmega
#> ℹ saving fit item: nlmixr2save
#> ℹ saving fit item: nlmixr2saveOrig
#> ℹ saving fit item: nobs
#> ℹ saving fit item: nsub
#> ℹ saving fit item: objDf
#> ℹ saving fit item: objective
#> ℹ saving fit item: ofvType
#> ℹ saving fit item: omega
#> ℹ saving fit item: origData
#> ℹ saving fit item: parFixed
#> ℹ saving fit item: parFixedDf
#> ℹ saving fit item: parHistData
#> ℹ saving fit item: phiM
#> ℹ saving fit item: qfirst
#> ℹ saving fit item: qw
#> ℹ saving fit item: qx
#> ℹ saving fit item: ranef
#> ℹ saving fit item: runInfo
#> ℹ saving fit item: saem0
#> ℹ saving fit item: saemControl
#> ℹ saving fit item: saemModel
#> ℹ saving fit item: scaleInfo
#> ℹ saving fit item: sessioninfo
#> ℹ saving fit item: shrink
#> ℹ saving fit item: table
#> ℹ saving fit item: time
#> ℹ saving fit item: tolFactor
#> ℹ saving fit item: ui
#> ℹ zipping fit files
#> ℹ removing unzipped fit files

# note this approach does not evaluate the first function only,
# everything else is evaluated so `suppressMessage(nlmixr(...))` would always run
# the model, never loading the cached model.

# The cached assignment operator works for generic functions too, but saves
# as a rds file.  For example,

sim := rxSolve(one.cmt, theo_sd) # First simulate
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments

sim := rxSolve(one.cmt, theo_sd) # Same arguments, load instead
#> ℹ sim.rds was not started with the same random state, removing and re-running
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments

sim := rxSolve(one.cmt, theo_sd, nStud=4) # Different arguments, resimulate
#> ℹ sim.rds does not match prior arguments, removing and re-running
#> ℹ parameter labels from comments are typically ignored in non-interactive mode
#> ℹ Need to run with the source intact to parse comments

Name		Name	Last commit message	Last commit date
Latest commit History 93 Commits
.github		.github
R		R
man		man
pkgdown/favicon		pkgdown/favicon
src		src
tests		tests
.Rbuildignore		.Rbuildignore
.gitignore		.gitignore
DESCRIPTION		DESCRIPTION
LICENSE		LICENSE
LICENSE.md		LICENSE.md
NAMESPACE		NAMESPACE
NEWS.md		NEWS.md
README.Rmd		README.Rmd
README.md		README.md
_pkgdown.yml		_pkgdown.yml
codecov.yml		codecov.yml

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

Installation

Example

About

Licenses found

Uh oh!

Releases

Packages

Uh oh!

Contributors

Uh oh!

Languages

Folders and files

Latest commit

History

Repository files navigation

Installation

Example

About

Topics

Resources

License

Licenses found

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages