在R语言中如何按多列分组计算分性别分年度的就业率

方案1：tidyverse生态实现（推荐，语法简洁易读） #

首先加载依赖包：

library(dplyr)
library(tidyr)

生成长格式统计结果 #

result_long <- my_data %>%
  # 按性别、年份分组
  group_by(Gender, Year) %>%
  summarise(
    # 组内总人数
    total = n(),
    # 组内就业人数
    employed_cnt = sum(Employment_status == "Employed"),
    # 计算就业率
    employment_rate = employed_cnt / total,
    .groups = "drop"
  )

转换为宽格式（和预期输出结构一致，年份为列、性别为行） #

result_wide <- result_long %>%
  pivot_wider(
    id_cols = Gender,
    names_from = Year,
    values_from = employment_rate
  )

方案2：Base R实现（无需安装加载第三方包） #

# 统计各分组总人数
total_cnt <- aggregate(ID ~ Gender + Year, my_data, length)
# 统计各分组就业人数
employed_cnt <- aggregate(Employment_status ~ Gender + Year, my_data, function(x) sum(x == "Employed"))
# 合并数据计算就业率
stat_res <- merge(total_cnt, employed_cnt, by = c("Gender", "Year"))
stat_res$employment_rate <- stat_res$Employment_status / stat_res$ID
# 转换为宽表
stat_wide <- reshape(
  stat_res, 
  idvar = "Gender", 
  timevar = "Year", 
  direction = "wide",
  drop = c("ID", "Employment_status")
)

注意事项 #

如果你的统计口径需要排除Other类的就业状态，可在统计前先做数据过滤：

# tidyverse 过滤写法
my_data <- filter(my_data, Employment_status %in% c("Employed", "Unemployed"))
# Base R 过滤写法
my_data <- my_data[my_data$Employment_status %in% c("Employed", "Unemployed"), ]

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