如何在R语言中用左侧非NA列值加_unclassified替换数据框NA值

可以实现需求，以下提供两种R语言的处理方法，均能将数据框中的NA值替换为左侧最近的非NA文本，并给替换后的内容添加_unclassified后缀。

方法一：使用dplyr + zoo包（tidyverse风格） #

该方法借助rowwise()按行处理数据，结合na.locf()快速填充左侧非NA值：

library(dplyr)
library(zoo)

# 构建示例数据框
feature <- c("1", "2", "3", "4", "5")
phylum <- c("Firmicutes", "Firmicutes", "Firmicutes", "Proteobacteria", "Firmicutes")
class <- c(NA, "Clostridia", "Clostridia", "Gammaproteobacteria", "Bacilli")
order <- c(NA, NA, "Oscillospirales", "Enterobacterales", "Staphylococcales")
family <- c(NA, NA, NA, "Enterobacteriaceae", "Staphylococcaceae")
genus <- c(NA, NA, NA, NA, "Staphylococcus")

df <- data.frame(feature, phylum, class, order, family, genus)

# 处理数据
df_processed <- df %>%
  rowwise() %>%
  mutate(
    taxa = c(phylum, class, order, family, genus),
    is_na = is.na(taxa),
    filled_taxa = na.locf(taxa, fromLast = FALSE),
    processed_taxa = ifelse(is_na, paste0(filled_taxa, "_unclassified"), filled_taxa),
    phylum = processed_taxa[1],
    class = processed_taxa[2],
    order = processed_taxa[3],
    family = processed_taxa[4],
    genus = processed_taxa[5]
  ) %>%
  select(-taxa, -is_na, -filled_taxa, -processed_taxa) %>%
  ungroup()

# 输出结果
print(df_processed)

方法二：Base R自定义函数（无需额外包） #

通过自定义遍历函数实现需求，无需加载第三方包：

# 构建示例数据框
feature <- c("1", "2", "3", "4", "5")
phylum <- c("Firmicutes", "Firmicutes", "Firmicutes", "Proteobacteria", "Firmicutes")
class <- c(NA, "Clostridia", "Clostridia", "Gammaproteobacteria", "Bacilli")
order <- c(NA, NA, "Oscillospirales", "Enterobacterales", "Staphylococcales")
family <- c(NA, NA, NA, "Enterobacteriaceae", "Staphylococcaceae")
genus <- c(NA, NA, NA, NA, "Staphylococcus")

df <- data.frame(feature, phylum, class, order, family, genus)

# 自定义填充函数：从左到右替换NA为最近非NA值并加后缀
fill_na_with_left <- function(vec) {
  filled_vec <- vec
  last_non_na <- NULL
  for (i in seq_along(vec)) {
    if (!is.na(vec[i])) {
      last_non_na <- vec[i]
    } else {
      filled_vec[i] <- paste0(last_non_na, "_unclassified")
    }
  }
  filled_vec
}

# 应用函数到每一行的分类列
df_processed_base <- df
df_processed_base[, -1] <- t(apply(df[, -1], 1, fill_na_with_left))

# 输出结果
print(df_processed_base)

两种方法处理后的结果均符合需求：

• feature 1的class、order、family、genus列均为Firmicutes_unclassified
• feature 2的order、family、genus列均为Clostridia_unclassified
• feature 3的family、genus列均为Oscillospirales_unclassified
• feature 4的genus列为Enterobacteriaceae_unclassified

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