The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
derive_shk()’s equation not rendering on the
pkgdown site by switching to katex math rendering. (#88)as_leveler() as the generic leveler
constructor, with an example of defining custom levels directly.
(#89)xtras_data(), a wrapper around
xpose::xpose_data() %>% as_xp_xtras() that silences the
rlang_warning-batched chatter
readr/dplyr can raise while parsing tables
with NaN/Inf or otherwise oddly formatted
values, without hiding warnings that indicate an actual read failure.
xp_from_bbr() now uses it internally. (#77)recalc_shk(), which recalculates eta shrinkage
directly from individual estimates rather than parsing NONMEM’s own
reported value, with a .etastype argument to exclude “true
zero” (as opposed to merely near-zero) etas from the calculation. Etas
are matched to their omega by column name where possible (eg
nlmixr2 models), falling back to NONMEM’s
ETA<k>/OMEGA(k,k) numbering convention.
(#79)shk variable type, plus
derive_shk()/backfill_shk() to populate it: a
per-individual eta shrinkage contribution diagnostic
(log((eta - mean(eta))^2)). (#78)shk_grid()/shk_vs_cov_grid()/shk_vs_contcov()/
shk_vs_catcov(), mirroring
eta_grid()/eta_vs_cov_grid()/
eta_vs_contcov()/eta_vs_catcov() for the new
shk diagnostic. (#78)list = FALSE to
eta_vs_contcov()/eta_vs_catcov()/
shk_vs_contcov()/shk_vs_catcov(), combining
the per-variable plots onto one shared, faceted plot instead of a list,
paginating at 9 panels per page. (#82)covvar to
eta_vs_contcov()/eta_vs_catcov()/
shk_vs_contcov()/shk_vs_catcov() to select
specific covariates, mirroring etavar/shkvar;
also added as a cols alias on
eta_vs_cov_grid()/shk_vs_cov_grid().
(#82)normalize_etas()/normalise_etas(),
which set a top-level xpdb$normalize_etas slot (like
$covs, see add_cov_association() – not an
xpdb$options entry, and not settable via
set_option()) dividing each eta by its omega-implied SD
(or, with .use_sd = TRUE, the empirical SD of its
individual estimates) in eta_grid()/
eta_vs_cov_grid()/eta_vs_contcov()/eta_vs_catcov()
only – the underlying data is never modified. (#81)nlmixr2_as_xtra() now reports the rejected object’s
class when xpose.nlmixr2::xpose_data_nlmixr2() doesn’t
recognize it as an nlmixr2 fit, instead of only the generic
upstream message.backfill_iofv() now gives a clear error when called
without xpdb instead of a cryptic cli
formatting failure. (#84)print.xpose_data()/print.xp_xtras()
erroring instead of printing whenever any xpdb$options
entry is a multi-element list (eg
default_labs/default_watermark with more than
one key set). (#81)xpdb$foo <- value/xpdb[["foo"]] <- value
(including inside patch_condn(), run automatically by
as_xpdb_x()) could silently strip the
xp_xtras/xpose_data classes on
ggplot2 versions before 4.0, due to a class-name collision
with ggplot2’s internal aes() mapping class.
(#74)conflicted is now an optional (Suggests) dependency
instead of a hard requirement, and is only consulted if the user has
loaded it themselves. Package startup no longer prints a “Registered S3
method overwritten” notice for print.xpose_plot, and the
reminder to attach xpose now also names any
load-order-sensitive bugfix at stake. (#72)irep() is no longer deprecated in favor of
xpose::irep() for xpose >= 0.5.0:
xpose reverted that upstream fix, so this package’s version
is needed again regardless of the installed xpose
version.catdv_vs_ipred() no longer triggers a spurious
ggplot2 lifecycle warning (incorrectly blaming
xpose) by mapping size to its line layer,
where it’s deprecated in favor of linewidth. (#75)set_var_types_x() no longer mistypes columns when a
tidyselect expression matches 10+ columns for one type
(e.g. eta = matches(...)), or when one type name is a
prefix of another (e.g. id/idv). (#76)xp_from_bbr() converts a ‘bbr’ model directly into an
xp_xtras object.add_cov_association(), prm_cov() family) with
xplot_forest() / cov_forest() visualization;
see the new parameter-associations vignette.cormat() visualizes the parameter
correlation/covariance matrix, for both NONMEM and nlmixr2 models.catdv_vs_occ() and catdv_vs_ipred() add
categorical DV diagnostic plots (observed-vs-predicted trend across
occasions, and binned calibration).left_join_x() backfills missing variables into an xpdb
via a coalescing left join.set_default_labs(), add_watermark(),
ggsave_xp()), configurable via
set_xtras_options() and auto-applied on print/save. See the
new plot-output vignette.patch_condn() corrects
‘xpose’s condition number calculation for multi-method NONMEM runs, applied automatically byas_xpdb_x()`.focus_function() / focus_qapply() now
support output-generating functions (plots, tables), not just
xpdb-transforming ones.logLik(), AIC(), and
BIC() methods for xpose_data and
xpose_set objects.ind_plots_sample() plots a representative (optionally
stratified) sample of individuals instead of the whole dataset.plot.xpose_data() runs a configurable batch of
diagnostic plots in one call.process_preset() family saves and replays a reusable
xpdb processing pipeline by name, optionally persisted across
sessions.set_dv_probs() no longer crashes when
.problem is omitted.conflicted preferences are now re-applied regardless of
package attach order, fixing spurious “found in 2 packages” errors.%p%.as_xpdb_x() / check_xpdb_x()
erroring on xpose_data objects with no table data.xplot_pairs() silently ignoring a custom
catcont_opts$other_fun, and removed an unreachable leftover
duplicate-axis-label block in roc_by_mod().nlmixr_example() / nlmixr2_example() to avoid
loading nlmixr2est when not installed.These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.