aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2022-10-26Don't test parhist and llhist on travis, docsJohannes Ranke18-73/+741
2022-10-26Reduced sfo model for multistart.saem.mkin testsJohannes Ranke6-45/+147
2022-10-26Remove kernel density estimate from llhistJohannes Ranke7-47/+112
2022-10-26Skip llhist and parhist tests as platform dependentJohannes Ranke1-0/+1
2022-10-26Try to make multistart (more) reproducibleJohannes Ranke4-152/+148
2022-10-26Tests pass locallyJohannes Ranke1-21/+19
2022-10-26Make multistart work with testthat scopesJohannes Ranke7-18/+301
Also skip long-running test with saem using the analytical from mkin
2022-10-25Test multistart methodJohannes Ranke2-0/+19
2022-10-25Round also for nlme print method...Johannes Ranke2-7/+8
2022-10-25Increase rounding in test for travisJohannes Ranke2-2/+2
As the "const" error model parameters is stored on log scale in nlme
2022-10-25Explanatory commentJohannes Ranke1-0/+1
2022-10-25Avoid another platform dependenceJohannes Ranke2-2/+7
Caused by the fact that printing error model parameters in nlme summaries do not honor the digits argument.
2022-10-25Change DFOP mixed model data in tests, updatesJohannes Ranke9-1282/+1298
2022-10-25Address platform dependencies that surfaced on travisJohannes Ranke2-5/+10
2022-10-25Another attempt to restore coverage on codecovJohannes Ranke1-2/+3
2022-10-25Install covr on travis to restore codecov coverageJohannes Ranke1-1/+1
2022-10-25Complete documentation and fix a bugJohannes Ranke11-29/+69
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
2022-10-21Explain chisq tests in anovaJohannes Ranke1-1/+2
Now that I understand how anova.lmer from which I ported the code produces mostly meaningful likelihood ratio tests in the table.
2022-10-21Adapt summary.saem.mmkin to covariate modelsJohannes Ranke1-7/+6
Dissipation times are calculated for covariate values of zero currently, which will usually not be very useful.
2022-10-21Add a simple anova method for model comparisonJohannes Ranke7-5/+175
Update docs
2022-10-21Make update.saem.mmkin work with objects from multistartJohannes Ranke1-6/+10
In particular, objects selected using best.multistart can now be updated, as the starting parameters are inserted into the call, without reference to 'start_parms' which will not be available outside the multistart method.
2022-10-20First working version setting up covariate modelsJohannes Ranke1-4/+53
2022-10-18Don't fail if one of the Hessians can't be computedJohannes Ranke1-1/+8
2022-10-17Fix selecting by log likelihood in parhistJohannes Ranke1-8/+21
2022-10-17SFORB in saem, update for mhmkin and multistartJohannes Ranke6-9/+94
2022-10-14update method for mhmkin objects, clarify commentJohannes Ranke2-2/+25
2022-10-14Select best fit from multistart, use in parhistJohannes Ranke6-31/+117
- Add 'best' and 'which.best' generics with methods for multistart objects - Per default, scale the parameters in parhist plots using the fit with the highest log likelihood.
2022-10-13Cope with failed FIM inversionsJohannes Ranke3-10/+29
2022-10-05Simplify the removal of random effectsJohannes Ranke5-29/+94
This is achieved by introducing the argument 'no_random_effect' to the saem function.
2022-10-04Add set_nd_nq to pkgdown topicsJohannes Ranke1-0/+1
2022-09-30Test the package with saemix development versionJohannes Ranke1-43/+12
2022-09-30Install development version of saemix on travisJohannes Ranke1-0/+1
I need this version for the parhist.multistart method, otherwise results for the two-component error model parameters can switch between positive and negative values.
2022-09-30r2u is now the default for r-ciJohannes Ranke1-1/+1
2022-09-29Add set_nd_nq previously in pfmJohannes Ranke7-18/+405
pfm depends on mkin anyways, so reexporting set_nd_nq and set_nd_nq_focus in pfm should provide reasonable continuity.
2022-09-29More adaptations to deal with failed saem fitsJohannes Ranke1-2/+8
2022-09-29Adapt to new format of failed saem.mmkin fitsJohannes Ranke2-8/+13
2022-09-28Update vignette and static docsJohannes Ranke8-21/+46
2022-09-28Fix handling of multistart fits with failuresJohannes Ranke13-104/+159
2022-09-26Make BIC.mhmkin work with failed fitsJohannes Ranke1-1/+4
2022-09-19Make the multistart method work on WindowsJohannes Ranke2-17/+43
2022-09-19Improve parhist and llhistJohannes Ranke13-59/+77
In particular, adapt the display of parameter boxplots for saem fits using mkin transformations to the way used for saem fits using saemix transformations, i.e. always show parameters on the natural scale, and normalised them by dividing by the median from the multiple runs.
2022-09-16Add llhist to the multistart vignetteJohannes Ranke5-1/+23
2022-09-16Improve multistart documentation, bugfixJohannes Ranke26-101/+1319
- Split out llhist and parhist documentation - Add example code for multistart - Create a multistart vignette, because the example code fails when run by pkgdown - Fix multistart for the case of mkin transformations in the saem fit
2022-09-16Tell travis to wait for coverage reportJohannes Ranke1-1/+1
2022-09-16Rerun test with saemix 3.1 from CRANJohannes Ranke3-14/+14
Apparently, the change I introduced into saemix saemixdevelopment/saemixextension#34 to fix error model parameters to positive values slightly influences the results. So this is now with the CRAN version to keep test results in sync with Travis
2022-09-16Improve docs and update pkgdownJohannes Ranke22-445/+560
2022-09-16Diagnostic plots for multistart methodJohannes Ranke13-111/+223
2022-09-14Remove purrr dependencyJohannes Ranke6-8/+11
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.
2022-09-13More work on multistartJohannes Ranke5-2/+18

Contact - Imprint