Merge branch 'master' into docs/data-functions-vignette-435

Author: ishaan-arora-1

.github/workflows/test-coverage.yaml

@@ -18,7 +18,7 @@ jobs:
       GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v6
 
       - uses: r-lib/actions/setup-r@v2
         with:

DESCRIPTION

@@ -26,12 +26,12 @@ Description: Plotting functions for posterior analysis, MCMC diagnostics,
 License: GPL (>= 3)
 URL: https://mc-stan.org/bayesplot/
 BugReports: https://github.com/stan-dev/bayesplot/issues/
-SystemRequirements: pandoc (>= 1.12.3), pandoc-citeproc
+SystemRequirements: pandoc (>= 1.12.3)
 Depends:
     R (>= 4.1.0)
 Imports:
     dplyr (>= 1.0.0),
-    ggplot2 (>= 3.4.0),
+    ggplot2 (>= 3.5.0),
     ggridges (>= 0.5.5),
     glue,
     lifecycle,
@@ -54,7 +54,7 @@ Suggests:
     rmarkdown (>= 1.0.0),
     rstan (>= 2.17.1),
     rstanarm (>= 2.17.4),
-    rstantools (>= 1.5.0),
+    rstantools (>= 2.0.0),
     scales,
     shinystan (>= 2.3.0),
     survival,

NAMESPACE

@@ -157,6 +157,7 @@ export(ppc_ribbon)
 export(ppc_ribbon_data)
 export(ppc_ribbon_grouped)
 export(ppc_rootogram)
+export(ppc_rootogram_grouped)
 export(ppc_scatter)
 export(ppc_scatter_avg)
 export(ppc_scatter_avg_data)
@@ -217,12 +218,11 @@ importFrom(dplyr,inner_join)
 importFrom(dplyr,left_join)
 importFrom(dplyr,mutate)
 importFrom(dplyr,n)
-importFrom(dplyr,one_of)
 importFrom(dplyr,pull)
 importFrom(dplyr,rename)
 importFrom(dplyr,select)
+importFrom(dplyr,slice_min)
 importFrom(dplyr,summarise)
-importFrom(dplyr,top_n)
 importFrom(dplyr,ungroup)
 importFrom(dplyr,vars)
 importFrom(ggplot2,"%+replace%")

NEWS.md

@@ -1,6 +1,12 @@
 # bayesplot (development version)
 
 * Added vignette sections demonstrating `*_data()` companion functions for building custom ggplot2 visualizations (#435)
+* Eliminate redundant data processing in `mcmc_areas_data()` by reusing the prepared MCMC array for both interval and density computation.
+* Validate equal chain lengths in `validate_df_with_chain()`, reject missing chain labels, and renumber data-frame chain labels internally when converting to arrays.
+* Added unit tests for previously untested edge cases in `param_range()`, `param_glue()`, and `tidyselect_parameters()` (no-match, partial-match, and negation behavior).
+* Bumped minimum version for `rstantools` from `>= 1.5.0` to `>= 2.0.0` .
+* Use `rlang::warn()` and `rlang::inform()` for selected PPC user messages instead of base `warning()` and `message()`.
+* Standardize input validation errors in `ppc_km_overlay()` and interpolation helpers to use `rlang::abort()` for consistent error handling.
 * Fix assignment-in-call bug in `mcmc_rank_ecdf()` (#).
 * Replaced deprecated `dplyr` and `tidyselect` functions (`top_n`, `one_of`, `group_indices`) with their modern equivalents to ensure future compatibility. (#431)
 * Documentation added for all exported `*_data()` functions (#209)
@@ -9,6 +15,10 @@
 * Fixed test in `test-ppc-distributions.R` that incorrectly used `ppc_dens()` instead of `ppd_dens()` when testing PPD functions
 * New functions `mcmc_dots` and `mcmc_dots_by_chain` for dot plots of MCMC draws by @behramulukir (#402)
 * Default to `quantiles=100` for all dot plots by @behramulukir (#402)
+* Use `"neff_ratio"` consistently in diagnostic color scale helpers to avoid relying on partial matching of `"neff"`.
+* Replace `expand = c(mult, add)` with `ggplot2::expansion()` helper in scale functions for consistency with ggplot2 >= 3.3.0 style.
+* Replace uses of `geom_bar(stat = "identity")` with the more idiomatic ggplot2 form `geom_col()` 
+* New function `ppc_rootogram_grouped` for grouped rootogram plots by @behramulukir and @jgabry (#419)
 
 # bayesplot 1.15.0
 

R/bayesplot-colors.R

@@ -189,7 +189,7 @@ color_scheme_view <- function(scheme = NULL) {
 
 #' @export
 print.bayesplot_scheme <- function(x, ...) {
-  tab <- data.frame(unlist(x, use.names = FALSE), stringsAsFactors = FALSE)
+  tab <- data.frame(unlist(x, use.names = FALSE))
   colnames(tab) <- attr(x, "scheme_name") %||% "hex_color"
   print(tab, ...)
 }
@@ -222,9 +222,8 @@ plot_scheme <- function(scheme = NULL) {
   )
 
   ggplot(color_data, aes(x = .data$name, y = .data$value, fill = .data$group)) +
-    geom_bar(
+    geom_col(
       width = .5,
-      stat = "identity",
       color = "white",
       linewidth = 0.1
     ) +

R/helpers-gg.R

@@ -37,10 +37,10 @@ modify_aes <- function(mapping, ...) {
 facet_wrap_parsed <- function(...) {
   facet_wrap(..., labeller = label_parsed)
 }
-dont_expand_y_axis <- function(expand = c(0,0)) {
+dont_expand_y_axis <- function(expand = expansion(0, 0)) {
   scale_y_continuous(expand = expand)
 }
-dont_expand_x_axis <- function(expand = c(0,0)) {
+dont_expand_x_axis <- function(expand = expansion(0, 0)) {
   scale_x_continuous(expand = expand)
 }
 dont_expand_axes <- function() {

R/helpers-mcmc.R

@@ -210,6 +210,13 @@ validate_df_with_chain <- function(x) {
     x$chain <- NULL
   }
   x$Chain <- as.integer(x$Chain)
+  if (anyNA(x$Chain)) {
+    abort("Chain values must not be NA.")
+  }
+  rows_per_chain <- table(x$Chain)
+  if (length(unique(rows_per_chain)) != 1) {
+    abort("All chains must have the same number of iterations.")
+  }
   x
 }
 
@@ -218,11 +225,14 @@ validate_df_with_chain <- function(x) {
 df_with_chain2array <- function(x) {
   x <- validate_df_with_chain(x)
   chain <- x$Chain
+  # Renumber arbitrary chain labels to the contiguous 1:N indices used internally.
+  chain <- match(chain, sort(unique(chain)))
   n_chain <- length(unique(chain))
   a <- x[, !colnames(x) %in% "Chain", drop = FALSE]
   parnames <- colnames(a)
   a <- as.matrix(a)
-  x <- array(NA, dim = c(ceiling(nrow(a) / n_chain), n_chain, ncol(a)))
+  n_iter <- nrow(a) %/% n_chain
+  x <- array(NA, dim = c(n_iter, n_chain, ncol(a)))
   for (j in seq_len(n_chain)) {
     x[, j, ] <- a[chain == j,, drop=FALSE]
   }

R/helpers-ppc.R

@@ -456,7 +456,7 @@ interpolate_gamma <- function(N, K, prob, L) {
 get_interpolation_values <- function(N, K, L, prob) {
   for (dim in c("L", "prob")) {
     if (all(get(dim) != .gamma_adj[, dim])) {
-      stop(paste(
+      abort(paste(
         "No precomputed values to interpolate from for '", dim, "' = ",
         get(dim),
         ".\n",
@@ -469,7 +469,7 @@ get_interpolation_values <- function(N, K, L, prob) {
   }
   vals <- .gamma_adj[.gamma_adj$L == L & .gamma_adj$prob == prob, ]
   if (N > max(vals$N)) {
-    stop(paste(
+    abort(paste(
       "No precomputed values to interpolate from for sample length of ",
       N,
       ".\n",
@@ -480,7 +480,7 @@ get_interpolation_values <- function(N, K, L, prob) {
     ))
   }
   if (N < min(vals$N)) {
-    stop(paste(
+    abort(paste(
       "No precomputed values to interpolate from for sample length of ",
       N,
       ".\n",
@@ -491,7 +491,7 @@ get_interpolation_values <- function(N, K, L, prob) {
     ))
   }
   if (K > max(vals[vals$N <= N, ]$K)) {
-    stop(paste(
+    abort(paste(
       "No precomputed values available for interpolation for 'K' = ",
       K,
       ".\n",
@@ -502,7 +502,7 @@ get_interpolation_values <- function(N, K, L, prob) {
     ))
   }
   if (K < min(vals[vals$N <= N, ]$K)) {
-    stop(paste(
+    abort(paste(
       "No precomputed values available for interpolation for 'K' = ",
       K,
       ".\n",

R/mcmc-diagnostics-nuts.R

@@ -183,7 +183,7 @@ mcmc_nuts_acceptance <-
         )
     }
     hists <- hists +
-      dont_expand_y_axis(c(0.005, 0)) +
+      dont_expand_y_axis(expansion(mult = 0.005, add = 0)) +
       facet_wrap(vars(.data$Parameter), scales = "free") +
       yaxis_text(FALSE) +
       yaxis_title(FALSE) +
@@ -483,8 +483,8 @@ mcmc_nuts_energy <-
       ) +
       scale_fill_manual("", values = fills, labels = aes_labs) +
       scale_color_manual("", values = clrs, labels = aes_labs) +
-      dont_expand_y_axis(c(0.005, 0)) +
-      scale_x_continuous(expand = c(0.2, 0)) +
+      dont_expand_y_axis(expansion(mult = 0.005, add = 0)) +
+      scale_x_continuous(expand = expansion(mult = 0.2, add = 0)) +
       labs(y = NULL, x = expression(E - bar(E))) +
       bayesplot_theme_get() +
       space_legend_keys()  +

R/mcmc-diagnostics.R

@@ -170,8 +170,8 @@ mcmc_rhat <- function(rhat, ..., size = NULL) {
     labs(y = NULL, x = expression(hat(R))) +
     scale_fill_diagnostic("rhat") +
     scale_color_diagnostic("rhat") +
-    scale_x_continuous(breaks = brks, expand = c(0, .01)) +
-    scale_y_discrete(expand = c(.025,0)) +
+    scale_x_continuous(breaks = brks, expand = expansion(mult = 0, add = 0.01)) +
+    scale_y_discrete(expand = expansion(mult = 0.025, add = 0)) +
     yaxis_title(FALSE) +
     yaxis_text(FALSE) +
     yaxis_ticks(FALSE)
@@ -199,7 +199,7 @@ mcmc_rhat_hist <- function(rhat, ..., binwidth = NULL, bins = NULL, breaks = NUL
     scale_color_diagnostic("rhat") +
     scale_fill_diagnostic("rhat") +
     labs(x = expression(hat(R)), y = NULL) +
-    dont_expand_y_axis(c(0.005, 0)) +
+    dont_expand_y_axis(expansion(mult = 0.005, add = 0)) +
     bayesplot_theme_get() +
     yaxis_title(FALSE) +
     yaxis_text(FALSE) +
@@ -255,14 +255,14 @@ mcmc_neff <- function(ratio, ..., size = NULL) {
       linetype = 2,
       linewidth = 0.25) +
     labs(y = NULL, x = expression(N[eff]/N)) +
-    scale_fill_diagnostic("neff") +
-    scale_color_diagnostic("neff") +
+    scale_fill_diagnostic("neff_ratio") +
+    scale_color_diagnostic("neff_ratio") +
     scale_x_continuous(
       breaks = breaks,
       # as.character truncates trailing zeroes, while ggplot default does not
       labels = as.character(breaks),
       limits = c(0, max(1, max_ratio) + 0.05),
-      expand = c(0, 0)) +
+      expand = expansion(0, 0)) +
     bayesplot_theme_get() +
     yaxis_text(FALSE) +
     yaxis_title(FALSE) +
@@ -287,10 +287,10 @@ mcmc_neff_hist <- function(ratio, ..., binwidth = NULL, bins = NULL, breaks = NU
       binwidth = binwidth,
       bins = bins,
       breaks = breaks) +
-    scale_color_diagnostic("neff") +
-    scale_fill_diagnostic("neff") +
+    scale_color_diagnostic("neff_ratio") +
+    scale_fill_diagnostic("neff_ratio") +
     labs(x = expression(N[eff]/N), y = NULL) +
-    dont_expand_y_axis(c(0.005, 0)) +
+    dont_expand_y_axis(expansion(mult = 0.005, add = 0)) +
     yaxis_title(FALSE) +
     yaxis_text(FALSE) +
     yaxis_ticks(FALSE) +
@@ -424,12 +424,12 @@ diagnostic_points <- function(size = NULL) {
 
 # Functions wrapping around scale_color_manual() and scale_fill_manual(), used to
 # color the intervals by rhat value
-scale_color_diagnostic <- function(diagnostic = c("rhat", "neff")) {
+scale_color_diagnostic <- function(diagnostic = c("rhat", "neff_ratio")) {
   d <- match.arg(diagnostic)
   diagnostic_color_scale(d, aesthetic = "color")
 }
 
-scale_fill_diagnostic <- function(diagnostic = c("rhat", "neff")) {
+scale_fill_diagnostic <- function(diagnostic = c("rhat", "neff_ratio")) {
   d <- match.arg(diagnostic)
   diagnostic_color_scale(d, aesthetic = "fill")
 }
@@ -541,9 +541,7 @@ drop_NAs_and_warn <- function(x) {
       bayesplot_theme_get()
     if (style == "bar") {
       graph <- graph +
-        geom_bar(
-          position = "identity",
-          stat = "identity",
+        geom_col(
           linewidth = 0.2,
           fill = get_color("l"),
           color = get_color("lh"),
@@ -574,7 +572,7 @@ drop_NAs_and_warn <- function(x) {
       scale_x_continuous(
         limits = c(-0.5, lags + 0.5),
         breaks = function(x) as.integer(pretty(x, n = 3)),
-        expand = c(0, 0)
+        expand = expansion(0, 0)
       ) +
       labs(x = "Lag", y = "Autocorrelation") +
       force_axes_in_facets()