基于Type与Sub_type配对批量绘制多列年度数据趋势图问题

需求说明 #

需使用R语言实现批量绘图：以year为X轴，遍历所有data_column_*列，同时按type与对应的sub_type（仅同type内的sub_type配对，不同type的同名sub_type不混合绘图）分组生成趋势图，例如Brand&Ambient、Brand&Food等组合。

示例数据 #

structure(list(year = c(2018, 2018, 2018, 2018, 2018, 2018, 2018, 
                        2018, 2018, 2019, 2019, 2019, 2019, 2019, 2019, 2019, 2019, 2020, 
                        2020, 2020), 
               type = structure(c(1L, 1L, 1L, 1L, 2L, 2L, 3L,3L, 3L, 1L, 1L, 1L, 1L, 2L, 3L, 3L, 3L, 1L, 1L, 1L), 
               levels = c("Brand","Shop", "Retail"), class = "factor"), 
               sub_type = structure(c(1L, 2L, 3L, 6L, 3L, 6L, 4L, 6L, 7L, 1L, 2L, 3L, 6L, 5L, 4L, 6L, 7L, 1L, 2L, 3L), 
              levels = c("Ambient", "Food", "Drink", "Grocery", "non-grocery", "non-drink","Food"), class = "factor"), 
              
data_column_1 = c(14, 11, 42, 68, 1, 0, 30, 20, 75, 23, 16, 0, 15, 0, 46, 26, 42, 19, 56, 0),
data_column_2 = c(NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_,  NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_,  NA_real_), 
data_column_3 = c(198, 527, 0, 224, 176, 0, 322, 2, 564, 973, 506, 0, 569, 0, 452, 422, 456, 720, 255, 0), 
data_column_4 = c(NA_real_,NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, 56, NA_real_, NA_real_, 62,NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_, NA_real_), 
data_column_5 = c(1600, 206, 150,36, 0, 0, 702, 263, 0, 1124, 215, 68,718, 0, 1938, 584, 0, 652, 181,67)), 

class = c("grouped_df", "tbl_df", "tbl", "data.frame"), 
row.names = c(NA, -20L), groups = structure(list(year = c(2018,2018, 2018, 2019, 2019, 2019, 2020), sigtype = structure(c(1L,2L, 3L, 1L, 2L, 3L, 1L), 
levels = c("Brand", "Shop","Retail"), class = "factor"), 
.rows = structure(list(1:4, 5:6,7:9, 10:13, 14L, 15:17, 18:20), 
ptype = integer(0), class = c("vctrs_list_of","vctrs_vctr", "list"))), 
class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, -7L), .drop = TRUE))

现有尝试代码 #

other.unknown.columns <- colnames(df)
other.unknown.columns <- other.unknown.columns[-c(1,2,3)] # remove year, type, sub_type

setwd("my/filepath")

df$type <- factor(df$sigtype)
df$subtype <- factor(df3$sub_type)

type.levels <- levels(df$type)
sub_type.levels <- levels(df$sub_type)

for (i in type.levels) {
    
  workingData <- subset(df, type == i) # this was an attempt to only have one type present, therefore to not iterate through unnecessary sub_types

    for (j in sub_type.levels) {
      
          for(k in other.unknown.columns){
            
           g <-  ggplot(workingData, aes(x=year, y=.data[[k]])) +
                geom_line() +
                geom_point() +
                ggtitle(heading)
           print(g)
      
      dev.off()         
            
      filename<-paste(i, "_", j, "_", k, ".png", sep="")
      png(filename = filename, width=720) 
      heading<-paste(i," sub_type ",j," - ", k)
      
      
  }
 }
}

dev.list()
dev.off()

遇到的问题 #

• 生成了所有type与sub_type的组合，而非仅同type内存在的sub_type配对；
• 绘图数据对应错误，出现单年度多个数据点的问题；
• dev.off使用易出现异常。

解决方案 #

核心修正点 #

1. 仅遍历当前type下存在的sub_type：不再使用全局的sub_type水平，而是对每个type的子集提取实际存在的sub_type值，避免生成无效组合。
2. 精准筛选分组数据：每个绘图数据集同时匹配type和sub_type，确保每个year对应唯一数据点。
3. 规范图形设备流程：先打开png设备，再绘图打印，最后关闭设备，避免顺序颠倒导致的异常。

修正后的代码 #

# 加载依赖包
library(ggplot2)

# 设置输出路径
setwd("my/filepath")

# 自动匹配所有data_column开头的列
data_cols <- colnames(df)[grepl("data_column_", colnames(df))]

# 获取所有唯一的type值
type_list <- unique(df$type)

# 遍历每个type
for (i in type_list) {
  # 获取当前type下实际存在的sub_type（去重）
  sub_type_list <- unique(df$sub_type[df$type == i])
  
  # 遍历当前type下的每个sub_type
  for (j in sub_type_list) {
    # 筛选当前type+sub_type的目标数据集
    plot_data <- df[df$type == i & df$sub_type == j, ]
    
    # 遍历每个数据列
    for (k in data_cols) {
      # 跳过全为NA的列，避免生成空图
      if (all(is.na(plot_data[[k]]))) next
      
      # 定义文件名和标题
      filename <- paste0(i, "_", j, "_", k, ".png")
      heading <- paste(i, "sub_type", j, "-", k)
      
      # 打开图形设备
      png(filename = filename, width = 720)
      
      # 绘制趋势图
      g <- ggplot(plot_data, aes(x = year, y = .data[[k]])) +
        geom_line() +
        geom_point() +
        ggtitle(heading) +
        theme_minimal() # 可选：美化图形主题
      
      print(g)
      
      # 关闭图形设备
      dev.off()
    }
  }
}

额外优化 #

• 自动识别data_column_开头的列，无需手动排除列，适配数据结构变化；
• 增加全NA列判断，跳过无有效数据的绘图任务；
• 修正原代码中的变量名错误（df3→df、sigtype→type）；
• 加入theme_minimal()美化图形，可根据需求替换其他主题。

内容的提问来源于stack exchange，提问作者MintySusan