Age | Commit message (Collapse) | Author | Files | Lines |
|
Also after the upgrade from buster to bullseye of my local system, some
test results for saemix have changed.
|
|
The errors in the example code were in the \dontrun sections, so they
were not caught by CRAN checks. In addition, the static help files
generated with pkgdown were cached, so I noticed the errors only
after completely regenerating the documentation for version 1.0.0.
|
|
- 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
|
|
Address release critical check and test issues
|
|
|
|
The experimental nlme version in my drat repository contains the
variance function structure varConstProp which makes it possible to use
the two-component error model in generalized nonlinear models using
nlme::gnls() and in mixed effects models using nlme::nlme().
|
|
|
|
|
|
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.
|
|
|
|
This is about twice as fast as deSolve compiled in the case of FOCUS D
|
|
https://github.com/r-lib/vdiffr/issues/86
|
|
Still in preparation for analytical solutions of coupled models
|
|
Preparing for symbolic solutions for more than one compound
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Merge DESCRIPTION manually to combine dependencies and rerun check to
update check.log
|
|
|
|
|
|
|
|
|
|
|
|
in the hope that this makes plotting cross-platform also for this error
model
|
|
|
|
Static documentation rebuilt by pkgdown
|
|
mkinfit objects now include an ll() function to calculate the
log-likelihood. Part of the code was refactored, hopefully making it
easier to read and maintain. IRLS is currently the default algorithm for
the error model "obs", for no particular reason. This may be subject
to change when I get around to investigate.
Slow tests are now in a separate subdirectory and will probably
only be run by my own Makefile target.
Formatting of test logs is improved.
Roundtripping error model parameters works with a precision of 10% when
we use lots of replicates in the synthetic data (see slow tests). This
is not new in this commit, but as I think it is reasonable this
closes #7.
|
|
|
|
Address winbuilder check problems, update check log, update of static docs
|
|
|
|
The new algorithm tries direct optimization of the likelihood, as well
as a three step procedure. In this way, we consistently get the
model with the highest likelihood for SFO, DFOP and HS for all 12
new test datasets.
|
|
Now we have a three stage fitting process for
nonconstant error models:
- Unweighted least squares
- Only optimize the error model
- Optimize both
Static documentation rebuilt by pkgdown
|
|
- Improve control of the number of cores
- Reduce the precision of the correlation matrix in the test summary
output, as the exact results are platform dependent
|
|
Static documentation rebuilt by pkgdown
|
|
Also:
- Change rounding in print.nafta
- Add dots argument to nafta()
- Use cores=1 in examples
- Restrict N in IORE model to values > 0
|
|
with respect to accuracy and robustness.
|
|
Static documentation rebuilt by pkgdown
|
|
- Rebuild static documentation
- Adapt test to new approach to two component error model
where the model is inadequate
|
|
Also improve the irls fitting of the error model and add a test
for FOCUS_2006_C where the second component of the error model is zero
|
|
|
|
Rename 'sigma_rl' to 'sigma_twocomp' as the Rocke and Lorenzato model assumes lognormal distribution for large y.
Rebuild static documentation.
|
|
|
|
Static documentation except articles rebuilt by pkgdown
|
|
Static documentation except articles rebuilt by pkgdown
|
|
|
|
|
|
|