pkgdown

.github/workflows/pkgdown.yaml

@@ -0,0 +1,50 @@
+# Workflow derived from https://github.com/r-lib/actions/tree/v2/examples
+# Need help debugging build failures? Start at https://github.com/r-lib/actions#where-to-find-help
+on:
+  push:
+    branches: [main, master]
+  pull_request:
+    branches: [main, master]
+  release:
+    types: [published]
+  workflow_dispatch:
+
+name: pkgdown
+
+permissions: read-all
+
+jobs:
+  pkgdown:
+    runs-on: ubuntu-latest
+    # Only restrict concurrency for non-PR jobs
+    concurrency:
+      group: pkgdown-${{ github.event_name != 'pull_request' || github.run_id }}
+    env:
+      GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: r-lib/actions/setup-pandoc@v2
+
+      - uses: r-lib/actions/setup-r@v2
+        with:
+          use-public-rspm: true
+
+      - uses: r-lib/actions/setup-r-dependencies@v2
+        with:
+          extra-packages: any::pkgdown, local::.
+          needs: website
+
+      - name: Build site
+        run: pkgdown::build_site_github_pages(new_process = FALSE, install = FALSE)
+        shell: Rscript {0}
+
+      - name: Deploy to GitHub pages 🚀
+        if: github.event_name != 'pull_request'
+        uses: JamesIves/github-pages-deploy-action@v4.5.0
+        with:
+          clean: false
+          branch: gh-pages
+          folder: docs

.gitignore

@@ -5,3 +5,4 @@ inst/doc
 .Rproj.user
 tests/testthat/Rplots.pdf
 docs
+.Rhistory

DESCRIPTION

@@ -73,4 +73,4 @@ Collate:
 RoxygenNote: 7.3.1
 Roxygen: list(markdown = TRUE)
 BugReports: https://github.com/optad/adoptr/issues
-URL: https://github.com/optad/adoptr
+URL: https://github.com/optad/adoptr, https://optad.github.io/adoptr/

_pkgdown.yml

@@ -1,7 +1,5 @@
 destination: docs
-
-url: https://optad.github.io/adoptr
-
+url: https://optad.github.io/adoptr/
 navbar:
   left:
   - icon: fa-home fa-lg
@@ -27,9 +25,6 @@ navbar:
   right:
   - icon: fa-github fa-lg
     href: https://github.com/optad/adoptr
-
 template:
-  params:
-    docsearch:
-      api_key: f5b8253ef131d8de31a2e0e075678ba8
-      index_name: adoptr
+  bootstrap: 5
+

vignettes/adoptr_jss.Rmd

@@ -271,8 +271,8 @@ $$
 of a two-stage design for given $n_1,c_1^f, c_1^e$ with respect to $n_2(\cdot)$
 for given power and type one error rate constraints.
 The function $c_2(\cdot)$, however, was not optimized.
-Instead, @JT2015 used a combination test approach to derive $c_2$ given
-$n_2(\cdot)$ and $n_1$ (cf. Equation \@ref(eq:inverse-normal)).
+Instead, @JT2015 used the inverse-normal combination test approach to derive $c_2$
+given $n_2(\cdot)$ and $n_1$.
 In @variational, the authors demonstrated that this restriction is not
 necessary and that the variational problem of deriving both functions $n_2(\cdot)$
 and $c_2(\cdot)$ given $n_1,c_1^f, c_1^e$ to minimize expected sample size can be
@@ -461,7 +461,7 @@ The package **adoptr** is based on **R**'s S4 class system.
 This allows to use multiple dispatch on the classes of multiple arguments 
 to a method.
 In this section, the central components of **adoptr** are described briefly.
-Figure \@ref(fig:class-diagram) gives a structural overview of the main
+The following figure gives a structural overview of the main
 classes in **adoptr**.
 
 ```{r class-diagram, out.width='100%', echo=FALSE, fig.cap='Overview of the most important classes and methods (in italic) in the R-package adoptr. A subclassing relationship is indicated by a connecting line to the  corresponding super class above it. The most important methods for each class are listed under the respective class name in italic font.'}
@@ -639,10 +639,9 @@ $x_1$ on the $x$-axis and the respective function on the $y$-axis.
 ```{r, standard-case, fig.height=2.25*1.25, fig.width=6*1.25, out.width='100%', fig.cap="Optimal sample size, critical value, and conditional power plotted against the interim test statistic (built-in plot method)."}
 plot(opt1$design, `Conditional power` = cp)
 ```
-Note the slightly bent shape of the $c_2(\cdot)$ function
-(cf. Figure \@ref(fig:standard-case), second plot).
+Note the slightly bent shape of the $c_2(\cdot)$ function.
 For two-stage designs based on the inverse-normal combination function, $c_2(\cdot)$
-would be linear by definition (cf. Equation \@ref(eq:inverse-normal)).
+would be linear by definition.
 Since the optimal shape of $c_2(\cdot)$ is not linear (but almost), inverse-normal
 combination methods are slightly less efficient (cf. @variational for a more
 detailed discussion of this issue).
@@ -755,7 +754,7 @@ evaluate(epower, opt3$design)
 ```
 
 The three optimal designs which have been computed so far are depicted
-in a joint plot (cf. Figure \@ref(fig:comparison)).
+in a joint plot.
 The design using the continuous prior requires higher sample sizes
 due to the higher uncertainty about $\theta$.
 The utility maximization approach results in similar shapes of $n(\cdot)$ and
@@ -877,8 +876,7 @@ opt4 <- minimize(ess, subject_to(toer <= 0.025, power >= 0.9, cp >= 0.8),
 Comparing the optimal design that has been computed here with the same constraints
 but without a conditional power constraint (cf. beginning of this chapter),
 the optimal design with the additional constraint requires larger sample sizes in
-regions where the conditional power would usually be below the given threshold
-(cf. Figure \@ref(fig:cp-constraint), first and third plot).
+regions where the conditional power would usually be below the given threshold.
 Overall, the additional constraint reduces the feasible solution space and
 consequently increases the expected sample size
 (`r if (buildvignette) evaluate(ess, opt4$design) %>% round(1)` with conditional power constraint
@@ -988,7 +986,7 @@ Note that it is also possible to 'un-fix' parameters again using the
 opt5 <- minimize(ess, subject_to(toer <= 0.025, power >= 0.9),
                  initial_design)
 ```
-Figure \@ref(fig:tunable) visually compares the original design with the new,
+The following figure visually compares the original design with the new,
 more restricted design.
 The designs are qualitatively similar, but fixing $n_1$ and $c_1^f$ does
 come at the price of slightly increased expected sample size