From 026fa0dd242c2d11637af2e2c88eed733f8c9f43 Mon Sep 17 00:00:00 2001
From: mattansb <35330040+mattansb@users.noreply.github.com>
Date: Mon, 29 Jun 2020 15:16:04 +0300
Subject: [PATCH] improve plot.bf_models code

---
 R/plot.bayesfactor_models.R        | 95 ++++++++++++++----------------
 man/plot.see_bayesfactor_models.Rd |  6 +-
 2 files changed, 50 insertions(+), 51 deletions(-)

diff --git a/R/plot.bayesfactor_models.R b/R/plot.bayesfactor_models.R
index 38f163c40..828d398b6 100644
--- a/R/plot.bayesfactor_models.R
+++ b/R/plot.bayesfactor_models.R
@@ -1,6 +1,7 @@
 #' Plot method for Bayes Factors for model comparison
 #'
 #' The \code{plot()} method for the \code{bayestestR::bayesfactor_models()} function.
+#' These plots visualize the \strong{posterior probabilities} of the compared models.
 #'
 #' @param n_pies Number of pies.
 #' @param value What value to display.
@@ -32,8 +33,11 @@
 #'
 #' result <- bayesfactor_models(lm1, lm2, lm3, denominator = lm0)
 #'
-#' plot(result, n_pies = "one", value = "probability") + theme_modern() +
+#' plot(result, n_pies = "one", value = "probability", sort = TRUE) +
 #'   scale_fill_pizza(reverse = TRUE)
+#'
+#' plot(result, n_pies = "many", value = "BF", log = TRUE) +
+#'   scale_fill_pizza(reverse = FALSE)
 #' @export
 plot.see_bayesfactor_models <-
   function(x,
@@ -57,72 +61,63 @@ plot.see_bayesfactor_models <-
     po_txt <- "NOTE: Slice sizes based on custom prior odds"
   }
 
-  if (isTRUE(sort)) {
-    one_pie_sort <- "PostProb"
-  } else {
-    one_pie_sort <- NULL
-  }
-
-  # Prep data and bar position:
+  ## Prep data
+  # One pie data
   one_pie_data <- as.data.frame(x)
-  one_pie_data$PostProb = (one_pie_data$BF / sum(one_pie_data$BF)) * priorOdds
+  one_pie_data$postOdds <- priorOdds * one_pie_data$BF
+  one_pie_data$PostProb <- (one_pie_data$postOdds / sum(one_pie_data$postOdds))
   if (isTRUE(sort)) one_pie_data <- one_pie_data[order(one_pie_data$PostProb, decreasing = TRUE), ]
-  one_pie_data$pos_txt <- sum(one_pie_data$PostProb) + one_pie_data$PostProb / 2 - cumsum(one_pie_data$PostProb)
-  one_pie_data$Model <- factor(one_pie_data$Model, levels = unique(one_pie_data$Model))
-
+  one_pie_data$Model <- factor(one_pie_data$Model, levels = one_pie_data$Model)
 
-
-  opd1 <- opd2 <- one_pie_data
-
-  opd1$Type <- denominator_name
-  opd1$BF <- one_pie_data$BF[denominator]
-  opd1$PostProb <- one_pie_data$PostProb[denominator]
-
-  opd2$Type <- one_pie_data$Model
-
-  many_pies_data <- rbind(opd1, opd2)
-  many_pies_data <- do.call(rbind, lapply(split(many_pies_data, many_pies_data$Model), function(.i) {
-    .i$pos_bar <- .i$PostProb / sum(.i$PostProb)
-    .i$pos_txt <- sum(.i$pos_bar) + .i$pos_bar / 2 - cumsum(.i$pos_bar)
-    .i[.i$Model != denominator_name, ]
-  }))
-  many_pies_data$Type <- factor(many_pies_data$Type, levels = unique(many_pies_data$Type))
+  # Two pie data
+  many_pies_data <- one_pie_data[one_pie_data$Model != denominator_name, ]
+  many_pies_data <- split(many_pies_data, many_pies_data$Model)
+  many_pies_data <- lapply(many_pies_data, function(m) {
+    m <- rbind(one_pie_data[one_pie_data$Model == denominator_name, ], m)
+    m$panel <- m$Model[2]
+    m$bar_pos <- m$PostProb / sum(m$PostProb)
+    m
+  })
+  many_pies_data <- many_pies_data[names(many_pies_data) != denominator_name]
+  many_pies_data <- do.call(rbind,many_pies_data)
+  many_pies_data$Model <- factor(many_pies_data$Model, levels = levels(one_pie_data$Model))
+  many_pies_data$panel <- factor(many_pies_data$panel, levels = levels(one_pie_data$Model))
+  many_pies_data$panel <- droplevels(many_pies_data$panel)
 
 
+  ## Labels
   if (value == "BF") {
     if (log) {
-      one_pie_data$label <- round(log(one_pie_data$BF), 2)
-      many_pies_data$label <- round(log(many_pies_data$BF), 2)
+      one_pie_data$label <- insight::format_value(log(one_pie_data$BF), 2)
+      many_pies_data$label <- insight::format_value(log(many_pies_data$BF), 2)
+      po_txt <- paste0(po_txt, "\nLabels are log(BF).")
     } else {
-      one_pie_data$label <- round(one_pie_data$BF, 2)
-      many_pies_data$label <- round(many_pies_data$BF, 2)
+      one_pie_data$label <- insight::format_value(one_pie_data$BF, 2)
+      many_pies_data$label <- insight::format_value(many_pies_data$BF, 2)
     }
   } else if (value == "probability") {
-    one_pie_data$label <- paste0(round(one_pie_data$PostProb * 100, 1),"%")
-    many_pies_data$label <- paste0(round(many_pies_data$PostProb * 100, 1),"%")
+    one_pie_data$label <- insight::format_value(one_pie_data$PostProb, 1, as_percent = TRUE)
+    many_pies_data$label <- insight::format_value(many_pies_data$PostProb, 1, as_percent = TRUE)
   } else {
     one_pie_data$label <- ""
     many_pies_data$label <- ""
   }
 
 
+  ## Plot
   if (n_pies == "one") {
-    ggplot(one_pie_data, aes(x = "", y = .data$PostProb, fill = .data$Model)) +
-      geom_bar(width = 1, stat = "identity", color = "white", size = .5) +
-      geom_text(aes(y = .data$pos_txt, label = .data$label), position = position_nudge(.1)) +
-      coord_polar("y", start = 0) +
-      scale_y_continuous(expand = c(0, 0)) +
-      labs(x = "", y = "", fill = "Model") +
-      theme_void() +
-      labs(caption = po_txt)
+    p <- ggplot(one_pie_data, aes(x = "", y = .data$PostProb, fill = .data$Model))
   } else {
-    ggplot(many_pies_data, aes(x = "", y = .data$pos_bar, fill = .data$Type)) +
-      geom_bar(width = 1, stat = "identity", color = "white", size = .5) +
-      geom_text(aes(y = .data$pos_txt, label = .data$label), position = position_nudge(.1)) +
-      coord_polar("y", start = 0) +
-      facet_wrap( ~ .data$Model) +
-      labs(x = "", y = "", fill = "Model") +
-      theme_void() +
-      labs(caption = po_txt)
+    p <- ggplot(many_pies_data, aes(x = "", y = .data$bar_pos, fill = .data$Model)) +
+      facet_wrap( ~ .data$panel)
   }
+
+  p +
+    geom_bar(width = 1, stat = "identity", color = "white", size = .5) +
+    geom_text(aes(label = .data$label), position = position_stack(vjust = 0.5)) +
+    coord_polar("y", start = 0) +
+    scale_y_continuous(expand = c(0, 0)) +
+    labs(x = "", y = "", fill = "Model") +
+    theme_void() +
+    labs(caption = po_txt)
 }
diff --git a/man/plot.see_bayesfactor_models.Rd b/man/plot.see_bayesfactor_models.Rd
index 628bc627d..1f09a1249 100644
--- a/man/plot.see_bayesfactor_models.Rd
+++ b/man/plot.see_bayesfactor_models.Rd
@@ -44,6 +44,7 @@ A ggplot2-object.
 }
 \description{
 The \code{plot()} method for the \code{bayestestR::bayesfactor_models()} function.
+These plots visualize the \strong{posterior probabilities} of the compared models.
 }
 \examples{
 library(bayestestR)
@@ -56,6 +57,9 @@ lm3 <- lm(qsec ~ drat + wt, data = mtcars)
 
 result <- bayesfactor_models(lm1, lm2, lm3, denominator = lm0)
 
-plot(result, n_pies = "one", value = "probability") + theme_modern() +
+plot(result, n_pies = "one", value = "probability", sort = TRUE) +
   scale_fill_pizza(reverse = TRUE)
+
+plot(result, n_pies = "many", value = "BF", log = TRUE) +
+  scale_fill_pizza(reverse = FALSE)
 }