scale_fill_viridis色条未显示最小断点0.02的问题求助

问题背景 #

我有一个名为sulfide的数据框，结构如下：

sulfide <- data.frame(SiteName = c("Site1", "Site1", "Site1", "Site1", "Site1", "Site1", "Site2", "Site2", "Site2", "Site2", "Site2", "Site2"), Activity = c("High", "High", "Medium", "Medium", "Low", "Low", "High", "High", "Medium", "Medium", "Low", "Low"), Depth = c("Shallow", "Deep", "Shallow", "Deep", "Shallow", "Deep", "Shallow", "Deep", "Shallow", "Deep", "Shallow", "Deep"), value = c(0.02, 0.02, 12.3, 8.6, 7.9, 9.7, 0.02, 0.02, NA, 8.6, 0.02, 1.0))

尝试绘制展示不同位点硫化物浓度及趋势的热力图，代码如下：

minimum <- 0.02

ggSulfide <- ggplot(sulfide, aes(x = Activity, y = Depth, fill = value)) + facet_grid(cols = vars(SiteName))
ggSulfide <- ggSulfide + geom_tile(color="white", size=0.1) + coord_equal()

ggSulfide <- ggSulfide + scale_fill_viridis(name = "Concentration\n(mM)", trans = "log", breaks = c(minimum, 0.2, 2, 10), labels = c(minimum, 0.2, 2, 10), limits = c(minimum, max(sulfide$value, na.rm = TRUE)), na.value = "white")

ggSulfide <- ggSulfide + geom_text(aes(label = value), size = geom.text.size)

ggSulfide

尽管已指定breaks、labels和绘图limits，色条仍未显示需要的最小值0.02（这是检测限，不能调整）。但将minimum改为0.01时，色条会显示该最小值。请问原因是什么？如何修复原色条？

原因分析 #

问题出在对数转换色条的刻度渲染逻辑上：

• 当你把色条范围limits的下限设为数据中的最小值0.02，同时使用log转换时，0.02是色条的边界点。ggplot2默认会自动隐藏和色条范围边界完全重合的断点，因为它默认边界是范围的起止，不需要额外标记。
• 当你把下限改成0.01时，0.02变成了色条范围内的一个普通点，而非边界，所以会被正常显示出来。

修复方法 #

方法1：强制显示边界刻度并取消自动扩展 #

在scale_fill_viridis中添加expand = c(0, 0)取消色条范围的自动扩展，同时通过guide_colorbar强制显示边界刻度：

minimum <- 0.02

ggSulfide <- ggplot(sulfide, aes(x = Activity, y = Depth, fill = value)) + 
  facet_grid(cols = vars(SiteName)) +
  geom_tile(color="white", size=0.1) + 
  coord_equal() +
  scale_fill_viridis(
    name = "Concentration\n(mM)", 
    trans = "log", 
    breaks = c(minimum, 0.2, 2, 10), 
    labels = c(minimum, 0.2, 2, 10), 
    limits = c(minimum, max(sulfide$value, na.rm = TRUE)), 
    na.value = "white",
    expand = c(0, 0), # 取消色条范围的自动扩展
    guide = guide_colorbar(show.limits = TRUE, ticks = TRUE) # 强制显示边界刻度
  ) +
  geom_text(aes(label = value), size = 3) # 替换未定义的geom.text.size为具体数值

ggSulfide

方法2：使用自定义对数转换域 #

使用scales::log_trans()指定转换的定义域，避免边界断点被忽略：

minimum <- 0.02

ggSulfide <- ggplot(sulfide, aes(x = Activity, y = Depth, fill = value)) + 
  facet_grid(cols = vars(SiteName)) +
  geom_tile(color="white", size=0.1) + 
  coord_equal() +
  scale_fill_viridis(
    name = "Concentration\n(mM)", 
    trans = scales::log_trans(domain = c(minimum, Inf)), 
    breaks = c(minimum, 0.2, 2, 10), 
    labels = c(minimum, 0.2, 2, 10), 
    limits = c(minimum, max(sulfide$value, na.rm = TRUE)), 
    na.value = "white",
    guide = guide_colorbar(show.limits = TRUE)
  ) +
  geom_text(aes(label = value), size = 3)

ggSulfide

内容的提问来源于stack exchange，提问作者as1234