如何为DataFrame的path列不同行子集应用不同时间缩放因子且不拆分数据？

问题背景 #

将多个CSV文件合并为一个DataFrame后，生成的path列对应每个CSV的ID，需要通过数学转换将其转为实验时间戳。合并后的combined_files tibble包含14列，path列为第一列。

数据规则：path1-7（对应前336行）与path8-100之间存在0.5小时的间隔，path1-7转换规则为(path-1)*0.5，path8-100则在此基础上加0.5小时偏移。

原代码错误分析 #

尝试直接对行子集做转换时出现语法错误，原代码片段：

files <- list.files(pattern = "*.csv")
files <- stringr::str_sort(files, numeric = TRUE)
combined_files <- readr::read_csv(files, id = "path") %>%
  mutate(path = str_remove(path, "cap_data_"))%>%
  mutate(path = str_remove(path, ".csv"))
combined_files$path <- as.numeric(combined_files$path)

# 错误的子集操作
(combined_files[combined_files[c(1:336), 1]] - 1)*0.5

报错内容：

Error in `vectbl_as_col_location()`:
! Can't subset columns with `combined_files[c(1:336), 1]`.
x `combined_files[c(1:336), 1]` must be logical, numeric, or character, not a <tbl_df/tbl/data.frame> object.

错误原因：combined_files[c(1:336), 1]返回的是一个tibble对象，而数据框子集索引需要的是逻辑向量、数值向量或字符向量，因此语法不合法。

现有临时解决方案 #

通过拆分数据框再合并的方式实现转换，代码如下：

combined_files_1st <- subset(combined_files, combined_files$path < 7.5 )
combined_files_2nd <- subset(combined_files, combined_files$path > 7)
combined_files_1st$path <- (combined_files_1st$path - 1)*0.5
combined_files_2nd$path <- (combined_files_2nd$path-1)*0.5 +0.5

combined_join <- rbind(combined_files_1st, combined_files_2nd)

无需拆分的优化方案 #

使用dplyr的case_when或ifelse函数，直接在原数据框中完成条件转换，无需拆分合并：

方案1：用case_when实现（逻辑清晰，易扩展） #

library(dplyr)
library(stringr)
library(readr)

# 合并CSV并一步完成path清洗与时间戳转换
combined_files <- read_csv(files, id = "path") %>%
  mutate(
    # 先清洗path列，转为数值型
    path = str_remove(path, "cap_data_") %>%
           str_remove(".csv") %>%
           as.numeric(),
    # 按规则转换为时间戳，直接替换原path列
    path = case_when(
      path <= 7 ~ (path - 1)*0.5,
      path >= 8 ~ (path - 1)*0.5 + 0.5,
      TRUE ~ NA_real_ # 处理异常值
    )
  )

方案2：用ifelse简化（逻辑简单时更简洁） #

combined_files <- combined_files %>%
  mutate(path = ifelse(path <=7, (path-1)*0.5, (path-1)*0.5 +0.5))

示例验证 #

针对提供的示例数据：

按照示例规则（path1-2转换为(path-1)*0.5，path3转换为(path-1)*0.5+1），用case_when实现：

sample_df <- tibble(
  path = c(1,1,1,2,2,2,3,3,3),
  p1 = c(10,20,30,10,20,30,10,20,30),
  p2 = c(20,30,54,20,30,54,20,30,54)
)

sample_df <- sample_df %>%
  mutate(path = case_when(
    path <=2 ~ (path-1)*0.5,
    path ==3 ~ (path-1)*0.5 +1,
    TRUE ~ NA_real_
  ))

转换后结果符合预期：

内容的提问来源于stack exchange，提问作者Kieran