R语言for循环中从不同数据框获取ggplot标题显示NA的解决求助

你正在使用R语言的ggplot2绘制批量图表，代码能按marker_id拆分数据绘图，但图表标题始终显示NA，无法调用marker_data数据框中explanation列的对应文本。

当前代码：

markers <- unique(patient_flow$marker_id)
marker_data <- ap_fcm_raw$markerdata

library(ggplot2)

for (i in markers){
p <- ggplot(data = subset(patient_flow, marker_id == i))+
    aes(x = as.numeric(pbd_cat), y = result, color = study)+
  ggtitle(marker_data$explanation[i])+
  geom_smooth()+
  geom_point()+
      labs(x = "Post Burn day")

ggsave(p, file=paste0("FCM_plots/plot_", i, ".png"))
}

marker_data的数据结构：

structure(list(marker_id = c("events_001", "ptot_002", "events_003", 
"ppar_004", "ptot_005", "events_006"), name = c("all_e", "all_e_t", 
"cell_e", "cell_p", "cell_t", "single_e"), cell_type = c("events", 
"events", "cells", "cells", "cells", "singlets"), gate = c("All Events\r\nEvents\r\n\r\n", 
"All Events\r\n%\r\nTotal\r\n\r\n", "Cells\r\nEvents\r\n\r\n", 
"Cells\r\n%\r\nParent\r\n\r\n", "Cells\r\n%\r\nTotal\r\n\r\n", 
"Singlets\r\nEvents\r\n\r\n"), explanation = c("All events in total", 
"Percentage of all total events", "Number of cells", "Percentage of Cells of Parent gate", 
"Percentage of cells of Total events", "Number of single events"
), panel = c("p2", "p2", "p2", "p2", "p2", "p2")), row.names = c(NA, 
-6L), class = c("tbl_df", "tbl", "data.frame"))

问题原因 #

问题出在索引匹配上：i是patient_flow$marker_id的字符型取值（比如"events_001"），但marker_data的行索引是默认的数字序列（1、2、3...），直接用marker_data$explanation[i]会尝试用字符值作为位置索引，导致无法匹配到对应的explanation文本，返回NA。

解决方案 #

方案1：将marker_data的行名设置为marker_id #

修改marker_data的行名，让行名与marker_id一致，这样就能直接用i索引对应的explanation：

markers <- unique(patient_flow$marker_id)
marker_data <- ap_fcm_raw$markerdata
# 设置行名为marker_id
rownames(marker_data) <- marker_data$marker_id

library(ggplot2)

for (i in markers){
p <- ggplot(data = subset(patient_flow, marker_id == i))+
    aes(x = as.numeric(pbd_cat), y = result, color = study)+
  # 现在可以直接通过i索引到对应的explanation
  ggtitle(marker_data$explanation[i])+
  geom_smooth()+
  geom_point()+
  labs(x = "Post Burn day")

ggsave(p, file=paste0("FCM_plots/plot_", i, ".png"))
}

方案2：用match函数获取对应位置索引 #

在循环中用match()找到当前i在marker_data$marker_id中的位置，再提取explanation：

markers <- unique(patient_flow$marker_id)
marker_data <- ap_fcm_raw$markerdata

library(ggplot2)

for (i in markers){
  # 找到当前marker_id对应的行位置
  idx <- match(i, marker_data$marker_id)
p <- ggplot(data = subset(patient_flow, marker_id == i))+
    aes(x = as.numeric(pbd_cat), y = result, color = study)+
  ggtitle(marker_data$explanation[idx])+
  geom_smooth()+
  geom_point()+
  labs(x = "Post Burn day")

ggsave(p, file=paste0("FCM_plots/plot_", i, ".png"))
}

方案3：提前合并数据框（推荐） #

将patient_flow和marker_data按marker_id合并，这样可以避免循环中的索引操作，代码更简洁且不易出错：

library(ggplot2)
library(dplyr)

# 合并两个数据框
combined_data <- patient_flow %>%
  left_join(marker_data, by = "marker_id")

# 按marker_id分组绘图
combined_data %>%
  group_by(marker_id) %>%
  group_walk(function(.data, .key){
    p <- ggplot(.data)+
      aes(x = as.numeric(pbd_cat), y = result, color = study)+
      ggtitle(.data$explanation[1])+ # 每组的explanation是相同的，取第一个即可
      geom_smooth()+
      geom_point()+
      labs(x = "Post Burn day")
    
    ggsave(p, file=paste0("FCM_plots/plot_", .key$marker_id, ".png"))
  })

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