Cannot set all violins to the same width in R Plotly - r

Question

I'm trying to reproduce in R Plotly a 2 categorical variables violin plot that works just fine in ggplot2. But when I set the widths of the individual violins to be the same, using scalemode = "width", as described in the reference (https://plotly.com/r/reference/violin/), it simply wont work. Instead, it shows the widths (violin maximum) proportional to the counts in each category.
Here is an example:
# Paths:
path_data = "data/"
path_lib = "renv/library/R-4.1/x86_64-pc-linux-gnu/"
# Packages:
require(dplyr, lib = path_lib)
require(readr, lib = path_lib)
require(RColorBrewer, lib = path_lib)
require(plotly, lib = path_lib)
# Dataset:
df = readr::read_csv(paste0(path_data, "nasa_exoplanets.csv")) %>%
as.data.frame()
attr(df, "spec") = NULL
df_varnames = readr::read_csv(paste0(path_data, "nasa_exoplanets_var_names.csv")) %>%
as.data.frame()
attr(df_varnames, "spec") = NULL
# Variables:
cat_var1 = "st_metratio"
cat_var2 = "disc_locale"
cat_var_name1 = (df_varnames %>%
dplyr::filter(var == cat_var1))$var_name
cat_var_name2 = (df_varnames %>%
dplyr::filter(var == cat_var2))$var_name
num_var = "sy_dist"
num_var_name = (df_varnames %>%
dplyr::filter(var == num_var))$var_name
# Adapt the data:
df_plot = df %>%
dplyr::select(cat_var1,
cat_var2,
num_var)
# Deal with NA:
df_plot[which(is.na(df_plot[, cat_var1])), cat_var1] = "NA"
df_plot[which(is.na(df_plot[, cat_var2])), cat_var2] = "NA"
df_plot = df_plot[which(!is.na(df_plot[, num_var])), ]
# Levels order:
sorted_levels1 = sort(unique(df_plot[, cat_var1]))
df_plot[, cat_var1] = factor(x = df_plot[, cat_var1],
levels = sorted_levels1)
sorted_levels2 = sort(unique(df_plot[, cat_var2]))
df_plot[, cat_var2] = factor(x = df_plot[, cat_var2],
levels = sorted_levels2)
# Plot:
my_palette = colorRampPalette(c("#111539", "#97A1D9"))
n_levels2 = length(unique(df_plot[, cat_var2]))
p = plot_ly(
data = df_plot,
type = "violin",
x = ~eval(parse(text = cat_var1)),
y = ~eval(parse(text = num_var)),
color = ~eval(parse(text = cat_var2)),
colors = my_palette(n_levels2),
spanmode = "hard",
alpha = 1,
box = list(visible = FALSE),
meanline = list(visible = FALSE),
points = FALSE,
scalemode = "width" ### this doesn't work ###
) %>%
layout(
xaxis = list(
title = paste0("<b>", cat_var_name1, "</b>"),
titlefont = list(size = 20),
tickfont = list(size = 18),
categoryorder = "array"
),
yaxis = list(
title = paste0("<b>", num_var_name, "</b>"),
titlefont = list(size = 20),
tickfont = list(size = 18),
type = "log"
),
margin = list(
l = 10,
r = 10,
t = 10,
b = 10
),
legend = list(
title = list(
text = paste0("<br><b>", cat_var_name2, "</b>"),
font = list(size = 18)
)
),
hoverlabel = list(font = list(size = 16)),
showlegend = TRUE,
violinmode = "group"
)
p
data file: https://github.com/rafael747cardoso/Data_Visualization_Gallery/blob/main/data/nasa_exoplanets.csv
How it should be, plotted in ggplot2:
How it is with R Plotly: