xpose.xtras

Introduction

This package adds some extra functionality and plots to the xpose framework. This includes some plots that have been missing in translation from xpose4, but also some useful features that truly extend the capabilities of what can be done with xpose.

There are a few bugfixes here and functionality which could easily be suggested as pull requests to the parent package. Given the size and broad use of xpose, it appears even minor pull requests take some time to implement. As such, this package implements those features directly and if at any point in the future these are added (perhaps in a better state) to the parent package, they will be deprecated if this package is in active use.

Load library(xpose) before library(xpose.xtras) so these bugfixes take effect via normal masking. If you use conflicted to manage namespace conflicts, load order stops mattering for them: xpose.xtras automatically registers its preferred side of these specific xpose/stats overrides. Conflicts against other, unrelated packages (e.g. another package’s own dplyr verb) aren’t affected by this and are still yours to resolve as usual (e.g. conflicted::conflicts_prefer()).

Installation

The package can be installed from CRAN:

install.packages("xpose.xtras")

The typical github installation will also work.

pak::pak("jprybylski/xpose.xtras")

The main github branch is reserved for CRAN release-ready versions. The dev branch is usually ahead if there is active development.

pak::pkg_install("jprybylski/xpose.xtras@dev")

Preview

The grandparent package, xpose4, used to have a nice collection of figures and documentation that is referred to as a “bestiarium”. The documentation site for this package serves as a complete bestiary, but see the uncommented examples below as a sort of menagerie. There is no assumption that these examples are self-explanatory, but hopefully users familiar with xpose will recognize the new (and renewed) tools made available by this package.

EBEs versus covariates

described <- xpdb_x %>%
  set_var_labels(AGE="Age", MED1 = "Digoxin", .problem = 1) %>%
  set_var_units(AGE="yrs")  %>%
  set_var_levels(SEX=lvl_sex(), MED1 = lvl_bin())
eta_vs_contcov(described,etavar=ETA1, quiet=TRUE)
#> `geom_smooth()` using formula = 'y ~ x'

eta_vs_catcov(described,etavar=ETA1, quiet=TRUE)

Shark plots

pheno_set %>%
  focus_qapply(backfill_iofv) %>%
  dofv_vs_id(run6, run9, quiet = TRUE)

Categorical DVs

pkpd_m3 %>%
  set_var_types(catdv=BLQ,dvprobs=LIKE) %>%
  set_dv_probs(1, 1~LIKE, .dv_var = BLQ) %>%
  set_var_levels(1, BLQ = lvl_bin()) %>%
  catdv_vs_dvprobs(quiet=TRUE)
#> `geom_smooth()` using method = 'gam' and formula = 'y ~ s(x, bs = "cs")'

nlmixr2_m3 %>%
  set_var_types(catdv=CENS,dvprobs=BLQLIKE) %>%
  set_dv_probs(1, 1~BLQLIKE, .dv_var = CENS) %>%
  set_var_levels(1, CENS = lvl_bin()) %>%
  roc_plot(quiet = TRUE)

Parameter correlation matrix

cormat(xpdb_x, quiet = TRUE)