R语言：获取与fixed_date匹配的列名失败，求高效实现方案

我有一个约66k行、150列的data.frame，所有列均为Date类型。示例数据如下：

structure(list(fixed_date = structure(c(19267, 19239, 19120, 
18035, 19253, 19211, 19120, 19778, 19120, 19211, 18224, 18224, 
19624, 19211, 19211, 19211, 18409, 18409, 18558, 19876, 19267
), class = "Date"), BEGIN1 = structure(c(19240, 19212, 19086, 
18008, 19240, 19177, 19086, 19751, 19086, 19177, 18197, 18197, 
19597, 19177, 19177, 18197, 18197, 18197, 18477, 18477, 19240
), class = "Date"), END1 = structure(c(19267, 19239, 19120, 18035, 
19253, 19211, 19120, 19778, 19120, 19211, 18224, 18224, 19624, 
19211, 19211, 18224, 18224, 18224, 18488, 18488, 19267), class = "Date"), 
    BEGIN2 = structure(c(NA, NA, 18043, NA, NA, NA, NA, NA, 18232, 
    NA, NA, NA, NA, NA, 18232, 18232, 18232, 18232, 18512, 19849, 
    18484), class = "Date"), END2 = structure(c(NA, NA, 18070, 
    NA, NA, NA, NA, NA, 18280, NA, NA, NA, NA, NA, 18280, 18280, 
    18280, 18280, 18523, 19876, 18495), class = "Date"), BEGIN3 = structure(c(NA, 
    NA, 18897, NA, NA, NA, NA, NA, 18323, NA, NA, NA, NA, NA, 
    18354, 18354, 18396, 18396, 18547, 19793, 18519), class = "Date"), 
    END3 = structure(c(NA, NA, 18924, NA, NA, NA, NA, NA, 18350, 
    NA, NA, NA, NA, NA, 18381, 18381, 18409, 18409, 18558, 19827, 
    18530), class = "Date"), BEGIN4 = structure(c(NA, NA, 18113, 
    NA, NA, NA, NA, NA, 18687, NA, NA, NA, NA, NA, 18722, 18372, 
    NA, NA, NA, NA, 18554), class = "Date"), END4 = structure(c(NA, 
    NA, 18140, NA, NA, NA, NA, NA, 18700, NA, NA, NA, NA, NA, 
    18749, 18399, NA, NA, NA, NA, 18565), class = "Date"), BEGIN5 = structure(c(NA, 
    NA, 18148, NA, NA, NA, NA, NA, 18407, NA, NA, NA, NA, NA, 
    18407, 18407, NA, NA, NA, NA, NA), class = "Date"), END5 = structure(c(NA, 
    NA, 18175, NA, NA, NA, NA, NA, 18434, NA, NA, NA, NA, NA, 
    18434, 18434, NA, NA, NA, NA, NA), class = "Date")), row.names = c(NA, 
-21L), class = c("tbl_df", "tbl", "data.frame"))

需求是新增一列，记录第2至11列中与第1列（fixed_date）值相等的列名。尝试以下代码无报错，但未生成目标列：

sample$match_date_<- apply(sample[,c(2:11)], 1, 
        function(x) names(which(x == sample[,1])))

想知道代码未生效的原因，同时需要更高效的实现方案。

原代码问题分析 #

• 匹配逻辑错误：apply逐行处理时，x是单行数据，而sample[,1]是整个fixed_date列向量，x == sample[,1]会将单行与整列做比较，无法得到逐行匹配的结果，最终which返回空值，导致赋值无效。
• 效率低下：apply是逐行循环处理，对66k行的大数据集来说速度较慢；且Date类型会被自动转为数值，虽然不影响比较，但增加了隐式转换的开销。

高效实现方案 #

方案1：向量化操作（最优性能） #

# 提取目标列的数值矩阵（Date转数值不影响比较逻辑）
target_cols <- as.matrix(sample[, -1])
# 生成逐行匹配的布尔矩阵
match_matrix <- target_cols == sample$fixed_date
# 提取每行匹配的列名并以逗号连接
sample$match_date_ <- apply(match_matrix, 1, function(x) {
  matched_names <- colnames(target_cols)[x]
  paste(matched_names, collapse = ", ")
})

• 优势：先通过向量化比较生成匹配矩阵，再逐行提取列名，比原始apply逐行处理快数倍，适合大尺寸数据集。

方案2：tidyverse风格（简洁易读） #

library(dplyr)
library(purrr)

sample <- sample %>%
  mutate(match_date_ = pmap_chr(select(., -fixed_date), 
                                ~ paste(names(c(...))[c(...) == fixed_date], collapse = ", ")))

• 优势：代码符合tidyverse语法，可读性强，处理66k行数据的效率优于原始apply。

方案3：data.table（超大数据集首选） #

如果数据量持续增长，data.table的内存效率和速度会更突出：

library(data.table)
setDT(sample)

sample[, match_date_ := {
  dt_subset <- .SD[, -"fixed_date"]
  apply(dt_subset == fixed_date, 1, function(x) paste(names(dt_subset)[x], collapse = ", "))
}]

• 优势：data.table的内存优化机制适合处理百万级以上的数据集，操作速度更快。

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