在R中基于Lookup Table生成殖民存在虚拟变量报错求助

错误原因 #

你遇到的missing value where TRUE/FALSE needed错误，核心有两个问题：

1. 殖民状态匹配错误：原代码中colonies_df[[paste0("col.", empire, ".", year)]]直接取整列数据，没有根据当前行的国家匹配colonies_df中对应的行，导致用整列值和单个组织的存在状态做逻辑判断，出现长度不匹配或NA。
2. 缺失值/无匹配处理缺失：当alldata中某行的国家没有对应的国家_年份列，或colonies_df中找不到对应国家的行时，会返回NA，导致if条件无法得到明确的TRUE/FALSE结果。

修复方案 #

方案1：修正原函数逻辑（保留循环写法） #

修改create_col_presence函数，确保匹配对应国家的殖民状态，并处理NA和无匹配的情况：

create_col_presence <- function(row, colonies_df, empire, year) {
  country <- row$country
  
  # 找到colonies_df中对应国家的行
  colony_match <- colonies_df[colonies_df$country == country, ]
  
  # 无匹配国家直接返回0
  if (nrow(colony_match) == 0) {
    return(0)
  }
  
  # 获取组织存在状态和殖民归属状态
  org_present <- row[[paste0(country, "_", year)]]
  colony_status <- colony_match[[paste0("col.", empire, ".", year)]]
  
  # 仅当两个状态都为1且无NA时返回1，否则返回0
  if (!is.na(org_present) && org_present == 1 && !is.na(colony_status) && colony_status == 1) {
    return(1)
  } else {
    return(0)
  }
}

# 循环生成新列（用vapply替代sapply，类型更稳定）
for (empire in c("belgium", "britain", "france", "germany", "italy", "netherlands", "portugal", "spain")) {
  for (year in c(1954, 1970, 1988, 2003, 2017)) {
    new_col_name <- paste0("colpresence_", empire, "_", year)
    alldata[[new_col_name]] <- vapply(1:nrow(alldata), function(i) {
      create_col_presence(alldata[i, ], colonies_df, empire, year)
    }, FUN.VALUE = integer(1))
  }
}

方案2：用tidyverse向量化处理（更高效推荐） #

通过转换为长格式数据匹配，避免逐行循环，效率更高且更易维护：

library(tidyverse)

# 1. 将alldata转为长格式：组织-国家-年份-存在状态
alldata_long <- alldata %>%
  pivot_longer(
    cols = matches("^[A-Za-z]+_[0-9]{4}$"),
    names_to = c("country", "year"),
    names_sep = "_",
    values_to = "org_present",
    values_drop_na = FALSE
  ) %>%
  mutate(year = as.integer(year))

# 2. 将colonies_df转为长格式：国家-年份-殖民帝国归属
colonies_long <- colonies_df %>%
  pivot_longer(
    cols = matches("^col\\.[a-z]+\\.[0-9]{4}$"),
    names_to = c(NA, "empire", "year"),
    names_sep = "\\.",
    values_to = "colony_of_empire",
    values_drop_na = FALSE
  ) %>%
  mutate(year = as.integer(year))

# 3. 合并数据并计算colpresence
combined <- alldata_long %>%
  left_join(colonies_long, by = c("country", "year")) %>%
  mutate(
    colpresence = ifelse(org_present == 1 & colony_of_empire == 1, 1, 0),
    colpresence_col = paste0("colpresence_", empire, "_", year)
  )

# 4. 将结果转回宽格式并合并回原alldata
colpresence_wide <- combined %>%
  select(-country, -year, -org_present, -empire, -colony_of_empire) %>%
  pivot_wider(
    names_from = colpresence_col,
    values_from = colpresence,
    values_fill = 0
  )

alldata <- alldata %>% bind_cols(colpresence_wide)

内容的提问来源于stack exchange，提问作者Agata T