geom_rain()绘图线条未连接数据点的问题解决咨询

问题说明 #

复制ggrain官网示例代码绘制rain图时，出现部分线条未准确连接对应数据点的问题（如左下角数据点无线条连接），同时触发Duplicated aesthetics after name standardisation: alpha警告。运行代码如下：

set.seed(42) # the magic number

iris_subset <- iris[iris$Species %in% c('versicolor', 'virginica'),]

iris.long <- cbind(rbind(iris_subset, iris_subset, iris_subset), 
                   data.frame(time = c(rep("t1", dim(iris_subset)[1]), rep("t2", dim(iris_subset)[1]), rep("t3", dim(iris_subset)[1])),
                              id = c(rep(1:dim(iris_subset)[1]), rep(1:dim(iris_subset)[1]), rep(1:dim(iris_subset)[1]))))

# adding .5 and some noise to the versicolor species in t2
iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t2"] <- iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t2"] + .5 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t2"]), sd = .2)
# adding .8 and some noise to the versicolor species in t3
iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t3"] <- iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t3"] + .8 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t3"]), sd = .2)

# now we subtract -.2 and some noise to the virginica species
iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t2"] <- iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t2"] - .2 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t2"]), sd = .2)

# now we subtract -.4 and some noise to the virginica species
iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t3"] <- iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t3"] - .4 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t3"]), sd = .2)

iris.long$Sepal.Width <- round(iris.long$Sepal.Width, 1) # rounding Sepal.Width so t2 data is on the same resolution
iris.long$time <- factor(iris.long$time, levels = c('t1', 't2', 't3'))

ggplot(iris.long[iris.long$time %in% c('t1', 't2'),], aes(time, Sepal.Width, fill = Species)) +
  geom_rain(alpha = .5, rain.side = 'f2x2', id.long.var = "id") +
  theme_classic() +
  scale_fill_manual(values=c("dodgerblue", "darkorange")) +
  guides(fill = 'none', color = 'none')
#> Warning: Duplicated aesthetics after name standardisation: alpha

问题原因 #

1. rain.side参数不匹配：'f2x2'是为4组（2×2）对比场景设计的布局逻辑，当前仅使用t1和t2两个时间点，搭配Species分组后实际是2组对比，用该参数会导致连线逻辑错误，部分数据点无法匹配。
2. 重复美学映射：geom_rain内部已处理alpha美学，外部单独传入alpha = .5会触发重复映射警告。

解决方案 #

修正步骤 #

• 替换rain.side为适配2组对比的参数，如'both'（双向连线）、'left'或'right'（单向连线）；
• 将alpha整合到填充颜色的透明度设置中，避免重复美学映射；
• 保留id.long.var = "id"确保每个个体的时间序列数据正确关联。

修正后代码 #

set.seed(42) # the magic number

iris_subset <- iris[iris$Species %in% c('versicolor', 'virginica'),]

iris.long <- cbind(rbind(iris_subset, iris_subset, iris_subset), 
                   data.frame(time = c(rep("t1", dim(iris_subset)[1]), rep("t2", dim(iris_subset)[1]), rep("t3", dim(iris_subset)[1])),
                              id = c(rep(1:dim(iris_subset)[1]), rep(1:dim(iris_subset)[1]), rep(1:dim(iris_subset)[1]))))

# adding .5 and some noise to the versicolor species in t2
iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t2"] <- iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t2"] + .5 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t2"]), sd = .2)
# adding .8 and some noise to the versicolor species in t3
iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t3"] <- iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t3"] + .8 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'versicolor' & iris.long$time == "t3"]), sd = .2)

# now we subtract -.2 and some noise to the virginica species
iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t2"] <- iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t2"] - .2 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t2"]), sd = .2)

# now we subtract -.4 and some noise to the virginica species
iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t3"] <- iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t3"] - .4 + rnorm(length(iris.long$Sepal.Width[iris.long$Species == 'virginica' & iris.long$time == "t3"]), sd = .2)

iris.long$Sepal.Width <- round(iris.long$Sepal.Width, 1) # rounding Sepal.Width so t2 data is on the same resolution
iris.long$time <- factor(iris.long$time, levels = c('t1', 't2', 't3'))

# 修正后的绘图代码
ggplot(iris.long[iris.long$time %in% c('t1', 't2'),], aes(time, Sepal.Width, fill = Species)) +
  geom_rain(rain.side = 'both', id.long.var = "id") +
  theme_classic() +
  scale_fill_manual(values=c(alpha("dodgerblue", 0.5), alpha("darkorange", 0.5))) +
  guides(fill = 'none', color = 'none')

效果说明 #

• rain.side = 'both'会让每个个体的t1和t2数据点双向连线，确保所有点都被正确连接；
• 通过alpha()函数设置填充色透明度，既保留半透明效果，又避免了重复美学映射的警告；
• id.long.var确保同一id的个体数据被正确关联，不会出现连线错位。

内容的提问来源于stack exchange，提问作者Goru