如何在R中按特定字符串“Spatially clustered”筛选DataFrame子集？

问题背景 #

需要从DataFrame中筛选出**任意一列包含字符串"Spatially clustered"**的行，且保留所有列，但以下两种尝试均未达到预期：

错误方法1（筛选列而非行）

moran_deviation_data_multiple_correction_1january_raw_pval_conclusion = data
moran_deviation_data_multiple_correction_1january_raw_pval_conclusion_spatially_clustered = select(moran_deviation_data_multiple_correction_1january_raw_pval_conclusion, matches("clustered"))

该代码用select(matches())是筛选列名含指定关键词的列，而非筛选行，完全偏离需求。

错误方法2（关键词错误）

moran_deviation_data_multiple_correction_1january_raw_pval_conclusion_spatially_clustered = moran_deviation_data_multiple_correction_1january_raw_pval_conclusion[apply(moran_deviation_data_multiple_correction_1january_raw_pval_conclusion,1, function(x) any(grepl("dispersed", x))), ]
moran_deviation_data_multiple_correction_1january_raw_pval_conclusion_spatially_clustered

该代码查找的是含"dispersed"的行，和需求的"Spatially clustered"关键词不符，自然得不到正确结果。

正确实现方法 #

方法1：Base R 原生实现

通过apply逐行检查，结合grepl匹配目标字符串：

# 定义要匹配的目标字符串
target_str <- "Spatially clustered"

# 筛选行：任意一列包含目标字符串则保留该行
filtered_df <- moran_deviation_data_multiple_correction_1january_raw_pval_conclusion[
  apply(moran_deviation_data_multiple_correction_1january_raw_pval_conclusion, 1, 
        function(row) any(grepl(target_str, row, fixed = TRUE))), 
  ]

# 查看筛选结果
filtered_df

• fixed = TRUE：开启精确字符串匹配，避免正则表达式干扰；如果需要模糊匹配可移除该参数
• apply(..., 1)：指定按行遍历DataFrame

方法2：Tidyverse 风格实现（dplyr + purrr）

更简洁的管道式写法：

library(dplyr)
library(purrr)

filtered_df <- moran_deviation_data_multiple_correction_1january_raw_pval_conclusion %>%
  filter(pmap_lgl(., ~any(grepl("Spatially clustered", c(...), fixed = TRUE))))

• pmap_lgl：逐行遍历所有列，返回逻辑值判断是否存在匹配项
• filter：根据逻辑值筛选符合条件的行

内容的提问来源于stack exchange，提问作者MK Huda