mkin/inst/rmarkdown, branch mainFitting kinetic models to chemical degradation data (also on github)https://cgit.jrwb.de/mkin/atom?h=main2025-09-12T20:16:13ZFacilitate centering covariates2025-09-12T20:16:13ZJohannes Rankejohannes.ranke@jrwb.de2025-09-12T20:16:13Zurn:sha1:17af45b984d348505527140470c7fff204748ea6
The motivation for this is to obtain confidence intervals for
covariate dependent parameters that are valid for a median/mean
value of the covariate.
Implementation by adding an argument to the 'saem' function, and
adapting the relevant functions working with 'saem' objects.
Vignettes, the template and tests were updated to use the new functionality.
Make parent model selection more robust2025-03-28T06:11:13ZJohannes Rankejohannes.ranke@jrwb.de2025-03-28T06:11:13Zurn:sha1:a789ea2ae74ae371476f0e1f035226d07a429a0b
In case one of the fits failed, the corresponding entry in `aic_parent`
is `NA`, so we need to tell the `min()` function to ignore `NA` values
instead of returning and empty result.
Fix an error in the template2023-10-09T19:29:05ZJohannes Rankejranke@uni-bremen.de2023-10-09T19:29:05Zurn:sha1:39dc3545f0ff6f561ba6193a96daa99cf06472a1
Thanks to Janina Wöltjen for pointing it out
Improve docs of multistart method2023-04-17T17:39:09ZJohannes Rankejranke@uni-bremen.de2023-04-17T17:39:09Zurn:sha1:004fce2520d6889d82226e21bc443426e81d93f2Make the template work on Windows (with RStudio)2023-04-16T11:31:14ZJohannes Rankejranke@uni-bremen.de2023-04-16T11:31:14Zurn:sha1:c99a12f1ccb03b46264c536bc9ded5d5de9f85df
It seems that we cannot use stored DLLs (for pathway fits) on clusters
when using Windows, where we use PSOCK clusters. Although it works on
winbuilder in the tests in tests/testthat/test_compiled.R.
Maybe it is something RStudio specific.
Increase test coverage2023-04-15T18:00:00ZJohannes Rankejranke@uni-bremen.de2023-04-15T18:00:00Zurn:sha1:f215080efba4097ccdaa2d1208edd36c26d7978a
Also, using mkin analytical solutions for more than one observed
variable is not supported (but could be if out_values would be
reordered).
Make predefined symbols safer2023-04-15T13:35:06ZJohannes Rankejranke@uni-bremen.de2023-04-15T13:35:06Zurn:sha1:b0b710ee9f9bb9bbe9708676d0c5822465e02203
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.
Neutral names for code chunks in the template2023-01-03T16:52:39ZJohannes Rankejranke@uni-bremen.de2023-01-03T16:52:39Zurn:sha1:fc692554ecdaff66548cc3e6d666e44b7aaaa9afImprove template text, use neutral name for dll_dir2023-01-03T16:31:47ZJohannes Rankejranke@uni-bremen.de2023-01-03T16:31:47Zurn:sha1:18a18ef46391c4b9fa3e5e1cb06937d90a83ba05Improved skeleton for hierarchical fits2023-01-03T08:31:48ZJohannes Rankejranke@uni-bremen.de2023-01-03T08:31:48Zurn:sha1:ed2ab7bba07e3cb036782227fe27943f4b5583fa
Now with working pathway fits using SFORB-SFO2 (only two parallel
metabolites instead of three) as the data for compound Ia was not
sufficient for a reliable fit.