R语言：如何通过for循环自动合并目标字符列表？

问题背景 #

尝试通过c()嵌套for循环生成目标字符向量，控制台能打印出正确内容，但最终cols变量始终为NULL，无法保存生成的结果。

原代码 #

cols <- 
 c(
   for(j in periods) {
     if(is.na(df[grep(j, df$Period),]$Scenario[[1]])) {
       print(j, collapse = ",")
   } else {
     for (k in ssp) {
       print(paste(j, k, sep="_", collapse = ","))
     }
   }  
 }
 )

控制台输出 #

# Console
[1] "1951-1980"
[1] "1981-2010"
[1] "2011-2040_ssp126"
[1] "2011-2040_ssp370"
[1] "2011-2040_ssp585"
[1] "2041-2070_ssp126"
[1] "2041-2070_ssp370"
[1] "2041-2070_ssp585"
[1] "2071-2100_ssp126"
[1] "2071-2100_ssp370"
[1] "2071-2100_ssp585"
> cols
NULL

期望生成的cols向量 #

cols <- c("1951-1980","1981-2010","2011-2040_ssp126","2011-2040_ssp370","2011-2040_ssp585","2041-2070_ssp126","2041-2070_ssp370","2041-2070_ssp585","2071-2100_ssp126","2071-2100_ssp370","2071-2100_ssp585")

附数据框 #

df <- structure(list(rowname = c("fr2041-2070_ssp126_5521", "fr2071-2100_ssp370_9962", 
"fr2071-2100_ssp585_6140", "fr2041-2070_ssp126_10362", "fr2071-2100_ssp585_2461", 
"fr2071-2100_ssp370_5668", "fr2011-2040_ssp585_5251", "fr2041-2070_ssp370_7120", 
"fr2071-2100_ssp126_7171", "fr2041-2070_ssp126_6092", "fr2071-2100_ssp370_12266", 
"fr2041-2070_ssp370_4224", "fr2041-2070_ssp370_3195", "fr2071-2100_ssp585_5245", 
"fr2041-2070_ssp126_5228", "fr2071-2100_ssp370_3492", "fr1951-1980_13223", 
"fr2071-2100_ssp585_4505", "fr2041-2070_ssp585_6574", "fr1981-2010_9649", 
"fr1981-2010_8049", "fr2071-2100_ssp370_1269", "fr2041-2070_ssp126_4494", 
"fr2071-2100_ssp370_2894", "fr2071-2100_ssp585_3951", "fr2041-2070_ssp126_3324", 
"fr2071-2100_ssp585_4510", "fr2041-2070_ssp370_7156", "fr2071-2100_ssp585_3029", 
"fr2041-2070_ssp585_2580"), Period = c("2041-2070", "2071-2100", 
"2071-2100", "2041-2070", "2071-2100", "2071-2100", "2011-2040", 
"2041-2070", "2071-2100", "2041-2070", "2071-2100", "2041-2070", 
"2041-2070", "2071-2100", "2041-2070", "2071-2100", "1951-1980", 
"2071-2100", "2041-2070", "1981-2010", "1981-2010", "2071-2100", 
"2041-2070", "2071-2100", "2071-2100", "2041-2070", "2071-2100", 
"2041-2070", "2071-2100", "2041-2070"), Scenario = c("ssp126", 
"ssp370", "ssp585", "ssp126", "ssp585", "ssp370", "ssp585", "ssp370", 
"ssp126", "ssp126", "ssp370", "ssp370", "ssp370", "ssp585", "ssp126", 
"ssp370", NA, "ssp585", "ssp585", NA, NA, "ssp370", "ssp126", 
"ssp370", "ssp585", "ssp126", "ssp585", "ssp370", "ssp585", "ssp585"
)), row.names = c(NA, -30L), class = c("data.table", "data.frame"
))

解决方案 #

核心原因 #

R的for循环本身不返回值，用c()包裹循环只能得到NULL；且print()仅输出内容到控制台，不会将值赋值给变量。需要先初始化空向量，再在循环中逐个添加元素。

修改后的代码 #

方法1：基础循环实现

# 定义periods和ssp（从数据中提取）
periods <- unique(df$Period)
ssp <- unique(df$Scenario[!is.na(df$Scenario)])

# 初始化空向量
cols <- c()

for(j in periods) {
  # 检查该Period对应的第一个Scenario是否为NA
  if(is.na(df[df$Period == j, ]$Scenario[[1]])) {
    cols <- c(cols, j)
  } else {
    for (k in ssp) {
      cols <- c(cols, paste(j, k, sep="_"))
    }
  }  
}

# 按期望顺序排序
target_order <- c("1951-1980","1981-2010","2011-2040_ssp126","2011-2040_ssp370","2011-2040_ssp585","2041-2070_ssp126","2041-2070_ssp370","2041-2070_ssp585","2071-2100_ssp126","2071-2100_ssp370","2071-2100_ssp585")
cols <- cols[match(target_order, cols)]

方法2：高效向量化实现（推荐）

避免循环，用向量化操作提升效率：

library(dplyr)

# 分组判断每个Period是否包含Scenario
period_data <- df %>%
  group_by(Period) %>%
  summarise(has_scenario = any(!is.na(Scenario))) %>%
  ungroup()

# 生成目标向量
cols <- c(
  # 无Scenario的Period直接保留
  period_data$Period[!period_data$has_scenario],
  # 有Scenario的Period与ssp组合
  outer(period_data$Period[period_data$has_scenario], ssp, paste, sep="_") %>% as.vector()
)

# 按目标顺序排序
target_order <- c("1951-1980","1981-2010","2011-2040_ssp126","2011-2040_ssp370","2011-2040_ssp585","2041-2070_ssp126","2041-2070_ssp370","2041-2070_ssp585","2071-2100_ssp126","2071-2100_ssp370","2071-2100_ssp585")
cols <- cols[match(target_order, cols)]

验证结果 #

运行代码后，cols将完全匹配期望的向量：

> cols
 [1] "1951-1980"          "1981-2010"          "2011-2040_ssp126"   "2011-2040_ssp370"   "2011-2040_ssp585"   "2041-2070_ssp126"  
 [7] "2041-2070_ssp370"   "2041-2070_ssp585"   "2071-2100_ssp126"   "2071-2100_ssp370"   "2071-2100_ssp585"

内容的提问来源于stack exchange，提问作者Tristan