aboutsummaryrefslogtreecommitdiff
path: root/R/saem.R
AgeCommit message (Collapse)AuthorFilesLines
2023-04-15Make predefined symbols saferJohannes Ranke1-2/+6
We still need to create a parallel processing cluster _after_ creating a compiled model that is moved to a user defined location, at least I did not find another way to make it work. This is not a problem with parallel processing without a cluster, which is not available on Windows.
2023-03-23Support covariates in endpoints()Johannes Ranke1-2/+2
2023-03-20Support covariates in parms and plot.saem.mmkinJohannes Ranke1-4/+31
2023-03-15Start parms and endpoints for covariatesJohannes Ranke1-0/+2
2023-02-17Finish adapting to upcoming deSolveJohannes Ranke1-7/+5
2023-02-13Merge branch 'main' into custom_lsoda_callJohannes Ranke1-1/+2
2022-11-18Pass error.init to saemix_model, show in parplotJohannes Ranke1-1/+2
Due to an oversight, error.init was not really passed to saemix_model in saem.mmkin. The new initial values were reverted to c(1, 1), in order to avoid changing the test results. Initial values for error model parameters are now shown in parplot.multistart.
2022-11-16We get about 25% performance gainJohannes Ranke1-6/+16
with the custom lsoda call, avoiding repeated getNativeSymbolInfo calls. It's just that we should not be calling foreign functions from different packages, because the may change without notice. Using getNativeSymbolInfo for "call_lsoda" avoids the CRAN note, and a similar call could probably be used for "unlock_solver", avoiding the NOTE in checks for cran, but we should not do this in a CRAN package.
2022-11-15Import saemix for easier profiling, update online docsJohannes Ranke1-1/+2
2022-11-14Automatic starting parameters for saem.mmkinJohannes Ranke1-1/+28
For the case of mkin transformations. This gives faster convergence, and appears to avoid problems with numeric ODE solutions
2022-11-08Same SFORB simplification for saem.mmkin fitsJohannes Ranke1-2/+2
2022-11-07Make print and summary for saem.mmkin more robustJohannes Ranke1-5/+10
If the likelihood computed by importance sampling, these methods failed. Now they report "Not available" or NA, respectively.
2022-11-04Never fail with errors in saem.mmkinJohannes Ranke1-6/+0
I think it is always preferable to be able to investigate the saemix object returned.
2022-10-31Possibility to override the error model in update.saemJohannes Ranke1-9/+10
2022-10-28Rename parhist to parplot and make it genericJohannes Ranke1-1/+1
That parhist name was not the brightest idea, as it does not show histograms.
2022-10-26Really test anova output, test covariatesJohannes Ranke1-0/+5
2022-10-25Complete documentation and fix a bugJohannes Ranke1-0/+1
The bug was introduced by the changes in summary.saem.mmkin.R and surfaced in the tests when using saemix transformations.
2022-10-25Revert accidental changeJohannes Ranke1-21/+27
2022-10-21Add a simple anova method for model comparisonJohannes Ranke1-4/+13
Update docs
2022-10-20First working version setting up covariate modelsJohannes Ranke1-4/+53
2022-10-17SFORB in saem, update for mhmkin and multistartJohannes Ranke1-1/+36
2022-10-14update method for mhmkin objects, clarify commentJohannes Ranke1-2/+2
2022-10-13Cope with failed FIM inversionsJohannes Ranke1-1/+2
2022-10-05Simplify the removal of random effectsJohannes Ranke1-4/+33
This is achieved by introducing the argument 'no_random_effect' to the saem function.
2022-09-28Fix handling of multistart fits with failuresJohannes Ranke1-48/+77
2022-09-16Diagnostic plots for multistart methodJohannes Ranke1-4/+4
2022-09-14Remove purrr dependencyJohannes Ranke1-0/+1
We now directly import from rlang and vctrs, which were indirect dependencies anyways. purrr::map_dfr is deprecated in the upcoming purrr 1.0, and depends on dplyr (since when?) which is only suggested by purrr. This would lead new installations of mkin to fail if dplyr is not installed as well.
2022-08-31Some more work on multistartJohannes Ranke1-4/+17
2022-08-10Batch processing for hierarchical fitsJohannes Ranke1-0/+34
- 'R/mhmkin.R': New method for performing multiple hierarchical mkin fits in one function call, optionally in parallel. - 'R/saem.R': 'logLik' and 'update' methods for 'saem.mmkin' objects. - 'R/illparms.R': Add methods for 'saem.mmkin' and 'mhmkin' objects. tests: Use 2 cores on travis, should work according to docs
2022-08-09Update docs for saemJohannes Ranke1-3/+3
2022-08-08Error on unsupported saemix transformationsJohannes Ranke1-2/+13
Also, support FOMC with saemix transformations
2022-08-08Fix fitting HS with saemix transformationsJohannes Ranke1-8/+19
2022-03-19Test saemix without special analytical solutionsJohannes Ranke1-12/+20
Also increase the performance a bit (from about 210 s to about 140 s in the case of DFOP with four chains and 300, 100 iterations).
2022-03-07Remove nlmixr interface for release of mkin 1.1.0Johannes Ranke1-1/+2
I am postponing my attempts to get the nlmixr interface to CRAN, given some problems with nlmixr using R-devel under Windows, see https://github.com/nlmixrdevelopment/nlmixr/issues/596 and https://github.com/r-hub/rhub/issues/512, which is fixed by the removal of nlmixr from the testsuite. For the tests to be more platform independent, the biphasic mixed effects models test dataset was defined in a way that fitting should be more robust (less ill-defined).
2022-03-01Make it possible to fix parameters in saem.mmkinJohannes Ranke1-2/+3
By actually passing dot arguments (that can include fixed.estim) to saemixModel, as advertised in the docs.
2022-02-09Adapt saemix interface to saemix 3.0 on CRANJohannes Ranke1-1/+8
2021-11-17Update required saemix version, update testsJohannes Ranke1-3/+2
2021-06-23Test log parameters by default when deriving saemix starting parametersJohannes Ranke1-1/+1
2021-06-09Rudimentary support for setting up nlmixr modelsJohannes Ranke1-0/+6
- All degradation models are specified as ODE models. This appears to be fast enough - Error models are being translated to nlmixr as close to the mkin error model as possible. When using the 'saem' backend, it appears not to be possible to use the same error model for more than one observed variable - No support yet for models with parallel formation of metabolites, where the ilr transformation is used in mkin per default - There is a bug in nlmixr which appears to be triggered if the data are not balanced, see nlmixrdevelopment/nlmixr#530 - There is a print and a plot method, the summary method is not finished
2021-03-20Fix mkin calculation of saemix residualsJohannes Ranke1-1/+1
2021-03-16Convenience option to set nbiter.saemixJohannes Ranke1-1/+5
2021-03-09Some improvements for borderline casesJohannes Ranke1-5/+26
- fit_with_errors for saem() - test_log_parms for mean_degparms() and saem()
2021-02-06Reintroduce interface to saemixJohannes Ranke1-0/+512
Also after the upgrade from buster to bullseye of my local system, some test results for saemix have changed.
2021-02-03Prepare for v1.0.0v1.0.0Johannes Ranke1-512/+0
- Improve authorship and copyright information - Prepare pkgdown config - Remove dependence on saemix as we need the development version which is not ready for CRAN - Temporarily remove saemix interface to check code coverage of the rest
2021-01-22We do not need to suppress saemix plots any moreJohannes Ranke1-12/+1
This makes fitting with saem within parallel::mclapply much faster and, surprisingly, much less hungry for RAM.
2020-12-19Possibility to override automatic starting parameters for saemJohannes Ranke1-4/+10
2020-12-15Remove leftoversJohannes Ranke1-3/+2
2020-12-09Make saem using mkinpredict work againJohannes Ranke1-46/+51
I threw out mclapply as it did not play well with the linear algebra routines used in the saemix code. Most of the change is actually indentation in the code creating the model function. But there is an important fix in mkinpredict which I had broken.
2020-12-09Add more tests and fix HS in saemJohannes Ranke1-2/+2
2020-12-08mixed.mmkin and test coverageJohannes Ranke1-1/+3

Contact - Imprint