R/dplyr：将两行数据转换为两列

嘿，这个需求很常见，用R的tidyverse工具链或者基础函数都能轻松实现，我给你写两种直接能用的方案：

方法1：用tidyr::pivot_wider（推荐，代码更直观） #

首先先构造你的原始数据框（方便你测试）：

# 构造原始数据
df <- data.frame(
  Word = c("shoe", "shoes", "toy", "toys", "key", "keys", "jazz"),
  Base = c("shoe", "shoe", "toy", "toy", "key", "key", "jazz"),
  Number = c(4834, 49955, 75465, 23556, 39485, 6546, 58765),
  Type = c("singular", "plural", "singular", "plural", "singular", "plural", "plural"),
  stringsAsFactors = FALSE
)

然后用pivot_wider做宽表转换，同时调整列名和顺序：

library(tidyr)
library(dplyr)

# 转换并整理数据
result <- df %>%
  # 按Base分组，把Type的不同值转成列
  pivot_wider(
    id_cols = Base,
    names_from = Type,
    values_from = c(Word, Number),
    # 缺失的单复数用NA填充
    values_fill = list(Word = NA_character_, Number = NA_integer_)
  ) %>%
  # 重命名列名匹配你的需求
  rename(
    Word_Sg = Word_singular,
    Word_Pl = Word_plural,
    Num_Singular = Number_singular,
    Num_Plural = Number_plural
  ) %>%
  # 调整列的顺序和你要的一致
  select(Word_Sg, Word_Pl, Base, Num_Singular, Num_Plural)

# 查看结果
print(result)

运行后输出就是你想要的结构：

Word_Sg Word_Pl Base Num_Singular Num_Plural
1    shoe   shoes shoe         4834      49955
2     toy    toys  toy        75465      23556
3     key    keys  key        39485       6546
4    NA     jazz jazz           NA      58765

方法2：Base R 实现（无需加载额外包） #

如果你不想加载tidyverse包，用基础R的reshape函数也能完成：

# Base R 转换
result_base <- reshape(
  df,
  idvar = "Base",
  timevar = "Type",
  direction = "wide",
  v.names = c("Word", "Number")
)

# 重命名列名
colnames(result_base) <- gsub("Word\\.(singular|plural)", "Word_\\1", colnames(result_base))
colnames(result_base) <- gsub("Number\\.(singular|plural)", "Num_\\1", colnames(result_base))
colnames(result_base) <- gsub("singular", "Sg", colnames(result_base))
colnames(result_base) <- gsub("plural", "Pl", colnames(result_base))

# 调整列顺序并替换空值为NA
result_base <- result_base[, c("Word_Sg", "Word_Pl", "Base", "Num_Sg", "Num_Pl")]
result_base[result_base == ""] <- NA

print(result_base)

这个方法的输出和第一种完全一致，适合环境受限不能加载包的场景。

内容的提问来源于stack exchange，提问作者rayne