如何在R的Plotly 3D散点图中实现部分类别点透明显示？

不用拆分多个轨迹，有两种简便方法实现需求：

方法一：给目标类别颜色添加透明度通道 #

Plotly支持带透明度的十六进制颜色（格式为#RRGGBBAA，最后两位是透明度，00全透明，FF完全不透明）。我们可以先把自定义颜色和品牌一一对应，再给指定品牌的颜色加上透明度：

mtcars$brand <- sapply(strsplit(rownames(mtcars), " "), "[[", 1)
mtcars$brand <- as.factor(mtcars$brand)

# 创建品牌-颜色的命名向量，确保颜色和品牌对应准确
brand_colors <- setNames(
  c("#ebac23", "#b80058", "#008cf9", "#006e00", "#00bbad",
    "#d163e6", "#b24502", "#ff9287", "#5954d6", "#00c6f8",
    "#878500", "#00a76c", "#bdbdbd", "#846b54", "#5F2F87",
    "#241346", "#B3EE3A", "#077187", "#FFE06A", "#73787E",
    "#A6CEE3", "#1F78B4"),
  levels(mtcars$brand)
)

# 给Cadillac和Ferrari的颜色添加透明度（这里用40表示半透明，可调整）
brand_colors[c("Cadillac", "Ferrari")] <- paste0(brand_colors[c("Cadillac", "Ferrari")], "40")

fig <- plot_ly(mtcars, x = ~wt, y = ~hp, z = ~qsec, 
               color = ~brand, colors = brand_colors,
               type = "scatter3d", mode = "markers")
fig

方法二：单独设置透明度列 #

新增一列控制每个点的透明度，再映射到Plotly的opacity参数：

mtcars$brand <- sapply(strsplit(rownames(mtcars), " "), "[[", 1)
mtcars$brand <- as.factor(mtcars$brand)

custom_colors <- c("#ebac23", "#b80058", "#008cf9", "#006e00", "#00bbad",
                   "#d163e6", "#b24502", "#ff9287", "#5954d6", "#00c6f8",
                   "#878500", "#00a76c", "#bdbdbd", "#846b54", "#5F2F87",
                   "#241346", "#B3EE3A", "#077187", "#FFE06A", "#73787E",
                   "#A6CEE3", "#1F78B4")

# 新增opacity列：Cadillac和Ferrari设为0.2（透明），其余为1（不透明）
mtcars$opacity <- ifelse(mtcars$brand %in% c("Cadillac", "Ferrari"), 0.2, 1)

fig <- plot_ly(mtcars, x = ~wt, y = ~hp, z = ~qsec, 
               color = ~brand, colors = custom_colors,
               opacity = ~opacity,  # 映射透明度列
               type = "scatter3d", mode = "markers")
fig

说明 #

• 方法一的好处是颜色和透明度绑定在一起，图例里的颜色也会显示对应透明度；
• 方法二更灵活，能单独调整透明度数值，不影响颜色本身。

两种方法都只需要一次绘制，不用逐个添加轨迹，完美解决你的需求。

内容的提问来源于stack exchange，提问作者Andrea