---
title: "Example scripts: Grid and Bayes selection"
description: "Local Grid/Bayes searches for binary, multiclass, pairwise and regression workflows with postprocessing."
output:
  rmarkdown::html_vignette:
    toc: true
    toc_depth: 2
vignette: >
  %\VignetteIndexEntry{Example scripts: Grid and Bayes selection}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

Select recipe, model and calibration settings using local searches, then assess
the complete procedure on independent data. Install the optional workflow
packages plus `tune`, `dials` and `yardstick`; postprocessing examples also use
the dependencies named in their scripts.

Start with `vignette("tuning", package = "vennabers")` for the
step-by-step guide. `?vennabers_examples` indexes all four example collections.

## Choose an example

- [Binary Grid selection](#binary-grid)
- [Shared multiclass Grid selection](#shared-multiclass-grid)
- [Shared multiclass Bayes selection](#shared-multiclass-bayes)
- [Pairwise OvO Grid selection](#pairwise-grid)
- [Pairwise OvO Bayes selection](#pairwise-bayes)
- [Regression Grid selection](#regression-grid)
- [Regression Bayes selection](#regression-bayes)
- [Grid selection with postprocessing](#postprocessing-grid)
- [Bayes selection with postprocessing](#postprocessing-bayes)

## Running an example

Each heading below names an installed script. To inspect and run one:

```{r run-example, eval=FALSE, purl=FALSE}
script <- system.file("examples", "binary-grid.R", package = "vennabers")
file.show(script)
source(script)
```

Replace the filename to run another script. The listings below show the complete
code without executing it during documentation builds. Sourcing a script runs
its fits and checks; resampling, tuning and composition matrices can take longer
than the introductory guides.

## Binary Grid selection {#binary-grid}

`binary-grid.R`: Compare joint and two-stage selection while preserving independent assessment data.

```{r binary-grid, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "binary-grid.R", package = "vennabers"))}
```

## Shared multiclass Grid selection {#shared-multiclass-grid}

`shared-multiclass-grid.R`: Select complete shared OvO/OvR procedures using joint or two-stage Grid searches.

```{r shared-multiclass-grid, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "shared-multiclass-grid.R", package = "vennabers"))}
```

## Shared multiclass Bayes selection {#shared-multiclass-bayes}

`shared-multiclass-bayes.R`: Select complete OvO/OvR class matrices with local stepwise Bayes searches.

```{r shared-multiclass-bayes, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "shared-multiclass-bayes.R", package = "vennabers"))}
```

## Pairwise OvO Grid selection {#pairwise-grid}

`pairwise-grid.R`: Search local class-pair grids and select using coupled class predictions.

```{r pairwise-grid, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "pairwise-grid.R", package = "vennabers"))}
```

## Pairwise OvO Bayes selection {#pairwise-bayes}

`pairwise-bayes.R`: Run local pair searches and assess the complete coupled predictor.

```{r pairwise-bayes, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "pairwise-bayes.R", package = "vennabers"))}
```

## Regression Grid selection {#regression-grid}

`regression-grid.R`: Select IVAR/CVAR model and moderation settings with joint or two-stage searches.

```{r regression-grid, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "regression-grid.R", package = "vennabers"))}
```

## Regression Bayes selection {#regression-bayes}

`regression-bayes.R`: Search IVAR/CVAR settings locally while keeping assessment observations untouched.

```{r regression-bayes, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "regression-bayes.R", package = "vennabers"))}
```

## Grid selection with postprocessing {#postprocessing-grid}

`postprocessing-grid.R`: Run separate B-local searches for classification and regression compositions, excluding P, C and assessment data.

```{r postprocessing-grid, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "postprocessing-grid.R", package = "vennabers"))}
```

## Bayes selection with postprocessing {#postprocessing-bayes}

`postprocessing-bayes.R`: Inspect local search budgets, complete candidate histories and recorded failures across learned compositions.

```{r postprocessing-bayes, eval=FALSE, purl=FALSE, code=readLines(system.file("examples", "postprocessing-bayes.R", package = "vennabers"))}
```