ggplot2交互式图表图例标签定制：仅保留极值位置自定义标签

Hey there! Let's solve this problem where you need to set custom legend labels only at the min and max values—perfect for when you don't know the value range upfront (like with interactive plots). Here's how to do it step by step:

Step 1: Dynamically Calculate Extremes #

First, we'll grab the minimum and maximum values directly from your dataset. This ensures it works no matter what your data looks like, even if it's generated interactively.

Step 2: Adjust the Fill Scale #

We'll modify scale_fill_viridis() to only display breaks at those extreme values, with your custom labels.

Here's the full working code based on your example:

set.seed(1990)
library(reshape2)
library(viridis)
library(ggplot2)

# Generate your sample data
df <- melt(outer(rnorm(100), rnorm(100)), varnames = c("X1", "X2"))

# Dynamically get min and max values from the 'value' column
min_val <- min(df$value)
max_val <- max(df$value)

# Build the plot with customized legend
p1 <- ggplot(df, aes(X1, X2)) + 
  geom_tile(aes(fill = value)) +
  scale_fill_viridis(
    # Only show legend breaks at min and max
    breaks = c(min_val, max_val),
    # Assign your custom labels to those breaks
    labels = c("Lowest value", "Highest value"),
    # Optional: Tweak legend appearance for cleaner look
    guide = guide_colorbar(
      ticks = FALSE,  # Hide tick marks if you don't need them
      frame.colour = "black"  # Add a frame around the color bar (optional)
    )
  ) +
  coord_equal()

p1

How It Works #

• Dynamic Extremes: By calculating min_val and max_val directly from df$value, you don't have to hardcode any values—this will adapt automatically to any dataset you use, including interactive ones.
• Custom Legend Breaks: The breaks parameter tells ggplot to only create legend markers at the min and max values. The labels parameter maps those breaks to your custom text.
• Optional Styling: The guide_colorbar() part lets you clean up the legend—hiding ticks removes unnecessary visual clutter, while adding a frame can make the legend easier to read.

If you're converting this to an interactive plot (like with plotly::ggplotly()), this legend configuration will carry over seamlessly since ggplotly inherits all ggplot2's legend settings.

内容的提问来源于stack exchange，提问作者Afiq Johari