Respect guide(aes = 'none') when constructing legend entries

NEWS.md

@@ -1,16 +1,20 @@
 # 4.10.0.9000
 
-## Improvements
+## New features
 
-* `ggplotly()` now supports the `{ggalluvial}` package. (#2061, @moutikabdessabour)
-* `ggplotly()` does not issue warnings with `options(warnPartialMatchArgs = TRUE)` any longer. (#2046, @bersbersbers)
+* `ggplotly()` now supports the `{ggalluvial}` package. (#2061, thanks @moutikabdessabour)
 
 ## Bug fixes
 
 * `ggplotly()` now converts `stat_ecdf()` properly. (#2065)
 * `ggplotly()` now correctly handles `geom_tile()` with no `fill` aesthetic. (#2063)
+* `ggplotly()` now respects `guide(aes = "none")` (e.g., `guide(fill = "none")`) when constructing legend entries. (#2067)
 * Fixed an issue with translating `GGally::ggcorr()` via `ggplotly()`. (#2012)
 
+## Improvements
+
+* `ggplotly()` does not issue warnings with `options(warnPartialMatchArgs = TRUE)` any longer. (#2046, thanks @bersbersbers)
+
 # 4.10.0
 
 ## Breaking changes in JavaScript API

R/layers2traces.R

@@ -102,12 +102,16 @@ layers2traces <- function(data, prestats_data, layout, p) {
   }
   # now to the actual layer -> trace conversion
   trace.list <- list()
+  aes_no_guide <- names(vapply(p$guides, identical, logical(1), "none"))
   for (i in seq_along(datz)) {
     d <- datz[[i]]
     # variables that produce multiple traces and deserve their own legend entries
-    split_legend <- paste0(names(discreteScales), "_plotlyDomain")
+    split_legend <- paste0(
+      setdiff(names(discreteScales), aes_no_guide), 
+      "_plotlyDomain"
+    )
     # add variable that produce multiple traces, but do _not_ deserve entries
-    split_by <- c(split_legend, "PANEL", "frame", split_on(d))
+    split_by <- c(split_legend, aes_no_guide, "PANEL", "frame", split_on(d))
     # ensure the factor level orders (which determines traces order)
     # matches the order of the domain values
     split_vars <- intersect(split_by, names(d))

tests/testthat/test-ggplot-legend.R

@@ -39,6 +39,26 @@ test_that("legend vanishes when theme(legend.position = 'none'')", {
   expect_identical(info$layout$showlegend, FALSE)
 })
 
+test_that("aesthetics can be discarded from legend with guide(aes = 'none')", {
+  df1 <- data.frame(
+    Date = seq(as.Date("2021-01-01"), as.Date("2021-01-10"), "days"),
+    Series = c(rep("SeriesA", 10), rep("SeriesB", 10)),
+    Values = rnorm(n = 20),
+    Mean = 0, V1 = 2, V2 = -2
+  )
+
+  p <- ggplot(df1, aes(x=Date, y=Values, color = Series, linetype = Series, shape = Series)) +
+    geom_line(aes(x = Date, y = Mean, color = "Mean", linetype = "Mean")) +
+    geom_line(aes(x = Date, y = V1, color = "QC", linetype = "QC")) +
+    geom_line(aes(x = Date, y = V2, color = "QC", linetype = "QC")) +
+    geom_line() + 
+    geom_point() + 
+    guides(shape = "none", linetype = "none")
+
+  expect_doppelganger(p, "guide-aes-none")
+})
+
+
 p <- ggplot(mtcars, aes(x = mpg, y = wt, color = factor(vs))) + 
   geom_point()