Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
The reason is that it is misleading in the case of saem.mmkin objects,
because convergence is not really checked there.
|
|
|
|
The bug was introduced by the changes in summary.saem.mmkin.R and
surfaced in the tests when using saemix transformations.
|
|
Update docs
|
|
|
|
- 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.
|
|
pfm depends on mkin anyways, so reexporting set_nd_nq and
set_nd_nq_focus in pfm should provide reasonable continuity.
|
|
|
|
|
|
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.
|
|
|
|
|
|
- '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
|
|
Also, add a method for gathering convergence information
and a method for gathering information on ill-defined parameters
|
|
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).
|
|
|
|
Also adapt summary.nlmixr.mmkin to correctly handle the way
formation fractions are translated to nlmixr
|
|
|
|
- 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
|
|
Also after the upgrade from buster to bullseye of my local system, some
test results for saemix have changed.
|
|
- 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
|
|
|
|
With automatic reloading in mkinfit and mkinpredict in case the
DLL is not loaded and the original DLL path has been cleaned up.
Depends on jranke/inline@974bdea04fcedfafaab231e6f359c88270b56cb9
See inline#13
|
|
|
|
- D24_2014 dataset on aerobic soil degradation of 2,4-D from the EU
assessment as mkindsg object with metadata
- f_time_norm_focus() to do time-step normalisation using the FOCUS
method
- focus_soil_moisture data with default moisture contents at pF1,
pF 2 and pF 2.5 for USDA soil types from FOCUS GW guidance
- Dataset generation scripts in inst/dataset_generation
- Depend on R >= 2.15.1 in order to facilitate the use of
utils::globalVariables()
|
|
|
|
With a plot method. The class mixed.mmkin is currently only a virtual
class created to unify the plotting method.
|
|
This avoids code duplication
|
|
|
|
Also make the endpoints function work for saem objects.
|
|
The reasons for this decision were
- Creating an saemix generic in the saemix package caused problems with
roxygen, because functions like saemix.plot.xy were documented in
their help files as S3 methods, although explicitly exported with
@export
- Creating an saemix generic in this package is possible, but would
make it necessary to load samix with exclude = "saemix" in order to
avoid overwriting the generic when loading saemix.
- The return object of such an saemix generic in this package cannot
be an S3 class with class attribute c("saemix.mmkin", "SaemixObject")
similar to nlme.mmkin, as saemix returns an S4 class.
- Extending the S4 class SaemixObject using simple inheritance to
a class SaemixMmkinObject with additional slots did not work
as expected. When the initialize method was left untouched, it
prevented creation of an SaemixMmkinObject even if it was based
on an initialised SaemixObject, as the initialize method seems
to always be called by new(). This could potentially be circumvented
by a coerce method. If an alternative initialize method was
used, an SaemixMmkinObject could be created. However, the methods
written for SaemixObjects only worked in some instances, either
because they checked for the class, and not for class inheritance
(like compare.saemix), or because the initialize method was called
for some reason. Therefore, the idea of creating a derived S4 class
was abandoned.
- A side effect of this decision is that the introduction of the saem
generic opens the possibility to use the same generic also for other
backends like nlmixr with the SAEM algorithm.
|
|
This commit also defined saemix.mmkin for mmkin row objects.
This works fine, but if we set the class of the returned object
to c("saemix.mmkin", "saemix"), it is not an S4 class any more
which make it impossible to use saemix functions on it.
|
|
Also, use logit transformation for g and for solitary formation
fractions, addressing #10.
|
|
Improve and update docs
|
|
Update docs
|
|
|
|
|
|
The method is no longer necessary, now that Bug 17761 is fixed upstream
https://bugs.r-project.org/bugzilla/show_bug.cgi?id=17761
|
|
This revealed a bug in the data returned in mkinfit$data in the case
of the d_3 algorithm, which also affected the residual plot - the
data from the direct fitting was not returned even if this was
the better method.
|
|
Also, use more intelligent starting values for the variance of the
random effects for saemix. While this does not appear to speed up
the convergence, it shows where this variance is greatly reduced
by using mixed-effects models as opposed to the separate independent
fits.
|
|
|
|
saemix_data depends on a development version of saemix, see
pull request saemixdevelopment/saemixextension#2
|
|
|
|
|
|
|
|
- mean_degparms() now optionally returns starting values for fixed and
random effects, which makes it possible to obtain acceptable fits
also in more difficult cases (with more parameters)
- Fix the anova method, as it is currently not enough to inherit from
lme: https://bugs.r-project.org/bugzilla/show_bug.cgi?id=17761
- Show fit information, and per default also errmin information
in plot.nlme.mmkin()
- Examples for nlme.mmkin: Decrease tolerance and increase the number of
iterations in the PNLS step in order to be able to fit FOMC-SFO and
DFOP-SFO
|
|
|
|
|
|
|