在R中绘制带绝对计数的PieDonut嵌套饼图报错求助

错误原因 #

showRatioPie参数仅接受逻辑值（TRUE/FALSE），用于控制是否显示饼图的比例标签，并非用来指定显示“绝对计数”的参数。传入字符串"absolute"会导致函数在执行逻辑判断时失败，从而抛出argument is not interpretable as logical错误。

解决方案 #

要同时显示饼图（外层）的百分比与绝对计数，需自定义标签函数结合统计数据实现。具体步骤如下：

1. 统计分类计数 #

先计算每个Class对应的观测数量，用于后续标签生成：

class_counts <- data[, .N, by = Class]
count_map <- setNames(class_counts$N, class_counts$Class)

2. 自定义标签函数 #

编写函数将百分比与绝对数合并为标签内容：

custom_pie_label <- function(a, r) {
  absolute <- round(a * nrow(data))  # 通过比例计算绝对数
  paste0(scales::percent(a), "\n(n=", absolute, ")")
}

3. 调用PieDonut生成图表 #

开启饼图比例标签显示，并传入自定义标签函数：

PieDonut(data, aes(Class, Gene), 
         showRatioPie = TRUE, 
         labelPie = custom_pie_label,
         showRatioDonut = function(a,r) scales::scientific(a))

完整可运行代码 #

library(ggpie)
library(data.table)

# 加载数据集
data <- structure(list(Class = c("ADTKD", "ADTKD", "Ciliopathy", "Cystic", 
"ADTKD", "Ciliopathy", "ADTKD", "Cystic", "ADTKD", "Systemic", 
"Glomerulopathy", "Collagenopathy", "Tubulopathy", "Systemic", 
"Ciliopathy", "Collagenopathy", "Glomerulopathy", "Cystic", "Tubulopathy", 
"Glomerulopathy", "Collagenopathy", "CAKUT", "Tubulopathy", "ADTKD", 
"Collagenopathy", "Collagenopathy", "CAKUT", "Ciliopathy", "Glomerulopathy", 
"Ciliopathy", "CAKUT", "Collagenopathy", "Cystic", "Collagenopathy", 
"Glomerulopathy", "ADTKD", "Collagenopathy", "CAKUT"), Gene = c("REN", 
"UMOD", "NPHP1", "OFD1", "UMOD", "NPHP1", "UMOD", "TSC2;PKD1", 
"UMOD", "EVC", "LMX1B", "COL4A4", "CLDN16", "GLA", "NPHP4", "COL4A4", 
"INF2", "PKD2", "CLDN16", "INF2", "COL4A4", "SALL1", "CLCN5", 
"UMOD", "COL4A4", "COL4A5", "PPP1R12A", "NPHP1", "TRIM8", "NPHP3", 
"17q12", "COL4A5", "PKD1", "COL4A3", "SMARCAL1", "UMOD", "COL4A5", 
"PAX2")), row.names = c(NA, -38L), class = c("data.table", "data.frame"
))

# 统计分类计数
class_counts <- data[, .N, by = Class]
count_map <- setNames(class_counts$N, class_counts$Class)

# 自定义饼图标签函数
custom_pie_label <- function(a, r) {
  absolute <- round(a * nrow(data))
  paste0(scales::percent(a), "\n(n=", absolute, ")")
}

# 绘制嵌套饼图
PieDonut(data, aes(Class, Gene), 
         showRatioPie = TRUE, 
         labelPie = custom_pie_label,
         showRatioDonut = function(a,r) scales::scientific(a))

关键说明 #

• showRatioPie=TRUE用于启用饼图标签，labelPie参数替换默认的百分比显示逻辑。
• 绝对数可通过a * nrow(data)计算（a为当前分类的比例值），也可直接使用预统计的count_map匹配分类名称获取。
• 环形图（Donut）的标签保留原代码的科学计数格式设置。

内容的提问来源于stack exchange，提问作者tacrolimus