Fixes & Test Refactoring

.Rbuildignore

@@ -13,3 +13,4 @@
 ^doc$
 ^Meta$
 ^tests
+^\.vscode$

.Rprofile

@@ -2,4 +2,5 @@ if(interactive()){
   library(devtools)
   library(testthat)
   library(vdiffr)
+  source("C:/R-Packages/BayesTools/tests/testthat/common-functions.R")
 }

.github/copilot-instructions.md

@@ -36,7 +36,7 @@ BayesTools is an R package for Bayesian analyses, JAGS model automation, and Bay
 
 ### Build & Test
 - **Install**: `devtools::install()` (Timeout: 5m+)
-- **Test**: `devtools::test()` (Timeout: 30m+)
+- **Test**: `devtools::test()` (Timeout: 5m+)
 - **Targeted Test**: `devtools::test(filter = 'priors')` (Recommended for dev)
 - **Check**: `rcmdcheck::rcmdcheck(args = c('--no-manual', '--as-cran'), error_on = 'never')` 
 - **Docs**: `devtools::document()` 
@@ -58,4 +58,11 @@ BayesTools is an R package for Bayesian analyses, JAGS model automation, and Bay
 ### Modifying JAGS Fitting
 1.  Edit `R/JAGS-fit.R` for general fitting logic or `R/JAGS-formula.R` for formula handling.
 2.  Verify with `devtools::test(filter = 'JAGS')`.
-3.  Ensure backward compatibility with `runjags` objects.
+3.  Ensure backward compatibility with `runjags` objects. 
+
+## Feature Addition Steps
+1. Create unit tests verifying the desired behavior.
+2. Implement the feature.
+3. Verify the tests pass.
+4. Update documentation as needed.
+5. Update NEWS.md with a summary of changes.

.github/instructions/r.instructions.md

@@ -0,0 +1,77 @@
+---
+description: 'R language and document formats (R, Rmd, Quarto): coding standards and Copilot guidance for idiomatic, safe, and consistent code generation.'
+applyTo: '**/*.R, **/*.r, **/*.Rmd, **/*.rmd, **/*.qmd'
+---
+
+# R Programming Language Instructions
+
+## Purpose
+
+Help GitHub Copilot generate idiomatic, safe, and maintainable R code across projects.
+
+## Core Conventions
+
+- **Match the project’s style.** Follow the style in the project.
+- **Prefer clear, vectorized code.** Keep functions small and avoid hidden side effects.
+- **Qualify non-base functions in examples/snippets**, e.g., `dplyr::mutate()`, `stringr::str_detect()`.
+- **Naming:** `lower_snake_case` for objects/files; use dots to dispatch different function types (and in S3 classes).
+- **Side effects:** Never call `setwd()`; prefer project-relative paths (e.g., `here::here()`).
+- **Validation:** Validate and constrain user inputs; use the predefined `check_bool()`, `check_char()`, `check_real()` ... functions.
+
+### Pipe Operators
+
+- **Never use pipe:** Always assign values using an arror `<-`
+
+## Performance Considerations
+
+- **Profiling:** Use `profvis::profvis()` to identify performance bottlenecks in your code. Profile before optimizing.
+- **Caching:** Use `memoise::memoise()` to cache expensive function results. Particularly useful for repeated API calls or complex computations.
+- **Vectorization:** Prefer vectorized operations over loops. Use `apply()` family for remaining iteration needs.
+
+## Tooling & Quality
+
+- **Pre-commit:** consider `precommit` hooks to lint/format automatically.
+- **Docs:** roxygen2 for exported functions (`@param`, `@return`, `@examples`).
+- **Tests:** prefer small, pure, composable functions that are easy to unit test.
+
+## Data Wrangling & I/O
+
+- **Data frames:** Use base `data.frame()`
+- **Iteration:** Prefer type-stable, vectorized patterns such as `vapply()` (for atomic outputs). Use `for` loops when when they improve clarity or performance.
+- **Strings & Dates:** Use clear base helpers (e.g., `nchar()`, `substr()`, `as.Date()` with explicit format).
+- **I/O:** prefer explicit, typed readers (e.g., `readr::read_csv()`); make parsing assumptions explicit.
+
+## Error Handling
+
+- Use `stop(..., .call = FALSE)` / `warning()`.
+- For recoverable operations:
+- Use `tryCatch()` in base R for fine-grained control.
+
+## Security Best Practices
+
+- **Command execution:** Prefer `processx::run()` or `sys::exec_wait()` over `system()`; validate and sanitize all arguments.
+- **File paths:** Normalize and sanitize user-provided paths (e.g., `fs::path_sanitize()`), and validate against allowlists.
+- **Credentials:** Never hardcode secrets. Use env vars (`Sys.getenv()`), config outside VCS, or `keyring`.
+
+## Copilot-Specific Guidance
+
+- Suggest vectorized solutions over loops when idiomatic.
+- Prefer small helper functions over long pipelines.
+- When multiple approaches are equivalent, prefer readability and type stability and explain the trade-offs.
+
+---
+
+## Minimal Examples
+
+```r
+scores <- data.frame(id = 1:5, x = c(1, 3, 2, 5, 4))
+safe_log <- function(x) tryCatch(log(x), error = function(e) NA_real_)
+scores$z <- vapply(scores$x, safe_log, numeric(1))
+
+# Example reusable helper with roxygen2 doc
+#' Compute the z-score of a numeric vector
+#' @param x A numeric vector
+#' @return Numeric vector of z-scores
+#' @examples z_score(c(1, 2, 3))
+z_score <- function(x) (x - mean(x, na.rm = TRUE)) / stats::sd(x, na.rm = TRUE)
+```

.github/instructions/tests.instructions.md

@@ -0,0 +1,105 @@
+---
+description: 'R unit tests: testing standards and Copilot guidance for interaction with unit tests.'
+applyTo: "**/tests/testthat/*.R"
+---
+
+# BayesTools Test Guidelines
+
+## Overview
+
+- Model fitting is centralized in `test-00-model-fits.R`; other tests load cached models
+- **testthat Edition 3** - do not use `context()` calls
+- Tests use `common-functions.R` for shared helpers
+
+## Test Caching (TDD Workflow)
+
+Model fitting is slow. The caching system lets you run the full suite once and reuse fits.
+
+### Environment Variables
+
+| Variable | Purpose | Default |
+|----------|---------|---------|
+| `BAYESTOOLS_TEST_FILES_DIR` | Cache directory location | `../temp/BayesTools_test_files` |
+| `BAYESTOOLS_TEST_SKIP_REFIT` | Skip fitting if cache exists | TRUE |
+
+### Recommended TDD Workflow
+
+```r
+# 1. Run full suite once to verify current code and populate cache (if missing) 
+devtools::test()
+
+# 2. Iterate on your feature - uses cached fits unless you are modifing model fitting!
+devtools::test(filter = "your-feature")
+
+# 3. Final verification (disable cache if fit / marglik code or its dependencies changed)
+clean_cached_fits()
+devtools::test()
+```
+
+### When to Clear Cache
+
+Clear with `clean_cached_fits()` when you modify:
+- `JAGS_fit()` or `JAGS_bridgesampling()` logic (or any of its dependencies)
+- Model definitions in `test-00-model-fits.R`
+
+## Key Rules
+
+### Model Fitting
+
+- **Only `test-00-model-fits.R`** fits models and computes marginal likelihoods
+- Other tests load with `readRDS(file.path(temp_fits_dir, "model_name.RDS"))`
+- Check `model_registry.RDS` for available models before creating new ones
+
+### File Naming
+
+| Pattern | Purpose |
+|---------|---------|
+| `test-{feature}.R` | Main tests |
+| `test-{feature}-input.R` | Input validation |
+| `test-{feature}-coverage.R` | Edge cases |
+
+### Skip Conditions
+
+| Condition | When to Use |
+|-----------|-------------|
+| `skip_if_no_fits()` | Test loads pre-fitted models |
+| `skip_if_not_installed("rjags")` | Test requires JAGS |
+| `skip_if_not_installed("bridgesampling")` | Test computes marginal likelihoods |
+| `skip_if_not_installed("vdiffr")` | Visual regression tests |
+
+### Helper Functions (common-functions.R)
+
+```r
+source(testthat::test_path("common-functions.R"))
+
+# Prior testing
+test_prior(prior)
+test_weightfunction(prior)
+test_orthonormal(prior)
+
+# Reference file testing
+test_reference_table(table, filename)
+test_reference_text(text, filename)
+
+# Skip/cache helpers
+skip_if_no_fits()
+skip_refit_if_cached(name)
+clean_cached_fits()
+```
+
+## AI Agent Protocol
+
+1. **Scan `test-00-model-fits.R` first** - understand available models
+2. **Reuse existing models** - don't create duplicates
+3. **Never fit models** outside `test-00-model-fits.R`
+4. **Never modify** `GENERATE_REFERENCE_FILES` flag (maintainer only)
+
+## Troubleshooting
+
+| Problem | Solution |
+|---------|----------|
+| "Pre-fitted models not available" | Run `devtools::test(filter = "00-model-fits")` |
+| Stale cache causing failures | `clean_cached_fits()` then rerun |
+| Tests pass locally, fail on CI | Clear cache, run full suite |
+
+````