在ggplot2中为不同图层设置不同调色板

需要绘制动物在活动场地中的运动轨迹，同时展示场地热图与按移动速度颜色编码的轨迹。已通过stat_density2d生成了BuGn色系的背景热图，但叠加geom_path后，轨迹的速度颜色区分度不足。

示例数据：

set.seed(100)
df <- data.frame(frame = seq(1,50,1),
             x = c(51.34564, 52.02254, 52.60115, 53.20906, 53.88028, 54.55151, 55.13741, 55.41307, 56.08957,
             56.68501, 57.28044, 57.62111, 58.13888, 58.38356, 58.38356, 58.38356, 58.17859, 57.52251,
             56.86643, 56.21035, 55.55427, 54.89819, 54.54640, 54.19461, 53.85371, 53.51281, 53.01192,
             52.51103, 51.82817, 51.14532, 50.46246, 49.82272, 49.18298, 48.38321, 47.58344, 46.78368,
             45.98391, 45.18414, 44.38437, 43.55679, 42.78776, 42.02832, 41.24797, 40.38921, 39.67093,
             38.78841, 38.07554, 37.04552, 36.21359, 35.42233),
             y = c(93.59669,  93.82952,  94.02557,  94.15208,  94.44488,  94.73768,  95.12280,  95.37992,
                   95.68827,  95.97229,  96.25632,  96.38436,  96.71693,  96.98220,  97.23499,  97.48778,
                   97.84184,  98.40484,  98.96784,  99.53085, 100.09385, 100.65685, 101.04628,  101.43570,
                   101.85187, 102.26803, 102.85582, 103.44360, 104.01978, 104.59596, 105.17213, 105.76429,
                   106.35644, 106.09024, 105.82403, 105.55783, 105.29163, 105.02542, 104.75922, 104.44431,
                   104.13191, 103.54044, 103.15739, 102.96974, 102.55943, 101.98391, 101.39337, 101.03815,
                   100.69386, 100.23876),
             speed = runif(50, min=0, max=5)) # 修正原数据speed长度与frame不匹配的问题

尝试的代码：

ggplot(df, aes(x,y)) + 
       xlab('X') + 
       ylab('Y') +
       stat_density2d(aes(fill=after_stat(density)), contour=FALSE, geom="raster", data=df) +
       scale_fill_distiller(palette="BuGn", direction=-1) +
       geom_path(data=df, aes(x=x, y=y, color=speed), size=1) +
       theme_classic()

要保留热图BuGn配色的同时提升轨迹速度颜色的区分度，只需为轨迹图层单独指定颜色刻度，以下是几种可行方案：

1. 使用Distiller内置渐变配色 #

选择与冷色调BuGn热图对比强烈的冷暖渐变配色，比如RdYlBu，通过scale_color_distiller设置：

ggplot(df, aes(x,y)) + 
  xlab('X') + 
  ylab('Y') +
  stat_density2d(aes(fill=after_stat(density)), contour=FALSE, geom="raster") +
  scale_fill_distiller(palette="BuGn", direction=-1) +
  geom_path(aes(color=speed), size=1) +
  # 为轨迹指定独立的颜色刻度，RdYlBu冷暖渐变与BuGn形成鲜明对比
  scale_color_distiller(palette="RdYlBu", direction=1, name="Speed") +
  theme_classic()

2. 使用Brewer离散配色（速度分箱展示） #

若希望将速度划分为多个区间展示，可配合cut()函数和scale_color_brewer的高饱和度配色：

ggplot(df, aes(x,y)) + 
  xlab('X') + 
  ylab('Y') +
  stat_density2d(aes(fill=after_stat(density)), contour=FALSE, geom="raster") +
  scale_fill_distiller(palette="BuGn", direction=-1) +
  # 将speed分为3个区间
  geom_path(aes(color=cut(speed, breaks=3)), size=1) +
  # Set1高饱和度配色，在绿色热图上清晰突出不同速度区间
  scale_color_brewer(palette="Set1", name="Speed Range") +
  theme_classic()

3. 自定义渐变配色 #

如果内置配色不符合需求，可通过scale_color_gradient创建完全自定义的渐变范围，确保与热图配色不冲突：

ggplot(df, aes(x,y)) + 
  xlab('X') + 
  ylab('Y') +
  stat_density2d(aes(fill=after_stat(density)), contour=FALSE, geom="raster") +
  scale_fill_distiller(palette="BuGn", direction=-1) +
  geom_path(aes(color=speed), size=1) +
  # 自定义从深蓝到深红的渐变，与绿色热图形成强烈对比
  scale_color_gradient(low="darkblue", high="darkred", name="Speed") +
  theme_classic()

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