这里有一个可能的解决方案，这个问题有点奇怪，因为你得到了stage和stock的重复值，而pivot_longer不知道如何自动折叠，相反，我们从一开始就删除重复值，然后它是一个单独的调用：

pivot_test %>%
  select(year, ends_with("Stk")) %>%
  pivot_longer(-year, names_to = "stock", values_to = "quantity") %>%
  mutate(stage=str_remove(stock, "Stk")) %>%
  select(year, stage, stock, quantity)

其返回

# A tibble: 18 × 4
    year stage  stock       quantity
   <int> <chr>  <chr>          <dbl>
 1  1991 tegg   teggStk   419648000 
 2  1991 tadult tadultStk     84009.
 3  1991 rec    recStk      1635970 
 4  1992 tegg   teggStk   411314000 
 5  1992 tadult tadultStk     58229.
 6  1992 rec    recStk      1269210 
 7  1993 tegg   teggStk   445784000 
 8  1993 tadult tadultStk     65226.
 9  1993 rec    recStk       927557 
10  1994 tegg   teggStk   434684000 
11  1994 tadult tadultStk     44372.
12  1994 rec    recStk      1312280 
13  1995 tegg   teggStk   364483000 
14  1995 tadult tadultStk     65555.
15  1995 rec    recStk      1387800 
16  1996 tegg   teggStk   686490000 
17  1996 tadult tadultStk     79611.
18  1996 rec    recStk      1772490

主要的挑战是获得正确的名称模式：

library(tidyr)
library(dplyr)
  
pivot_test %>% 
    pivot_longer(cols = -year, 
                 names_to = c("stage", "stock"), 
                 names_pattern = "(.*)(Stk)", 
                 values_to = "quantity") %>% 
  na.omit() %>% 
  mutate(stock = paste0(stage, stock))

year stage  stock       quantity
   <int> <chr>  <chr>          <dbl>
 1  1991 tegg   teggStk   419648000 
 2  1991 tadult tadultStk     84009.
 3  1991 rec    recStk      1635970 
 4  1992 tegg   teggStk   411314000 
 5  1992 tadult tadultStk     58229.
 6  1992 rec    recStk      1269210 
 7  1993 tegg   teggStk   445784000 
 8  1993 tadult tadultStk     65226.
 9  1993 rec    recStk       927557 
10  1994 tegg   teggStk   434684000 
11  1994 tadult tadultStk     44372.
12  1994 rec    recStk      1312280 
13  1995 tegg   teggStk   364483000 
14  1995 tadult tadultStk     65555.
15  1995 rec    recStk      1387800 
16  1996 tegg   teggStk   686490000 
17  1996 tadult tadultStk     79611.
18  1996 rec    recStk      1772490

为什么不直接将对tidyr::pivot_longer()的“stock”调用转换为对tidyr::pivot_longer()的“stage”调用，反之亦然呢？

pivot_test <- structure(list(
      year = 1991:1996,
      rec = c(1635970, 1269210, 927557, 1312280, 1387800, 1772490),
      teggStk = c(419648000, 411314000, 445784000, 434684000, 364483000, 686490000),
      tadultStk = c(84009.2, 58228.6, 65226.2, 44371.9, 65555.1, 79610.6),
      tadult = c(84009.2, 58228.6, 65226.2, 44371.9, 65555.1, 79610.6),
      teggs = c(419648000, 411314000, 445784000, 434684000, 364483000, 686490000),
      recStk = c(1635970, 1269210, 927557, 1312280, 1387800, 1772490)),
      row.names = c(NA, 6L), class = 'data.frame')
    
    annual21atest <- 
      tidyr::pivot_longer(pivot_test,
                          -c('year', 'recStk', 'teggStk', 'tadultStk'),
                          names_to= 'stage',
                          values_to = 'quantity')|>
      tidyr::pivot_longer(-c('year', 'stage'),
                          names_to= 'stock',
                          values_to = 'quantity')|>
      as.data.frame()

结果就是您所要求的：

year  stage     stock    quantity
    1  1991    rec   teggStk 419648000.0
    2  1991    rec tadultStk     84009.2
    3  1991    rec    recStk   1635970.0
    4  1991    rec  quantity   1635970.0
    5  1991 tadult   teggStk 419648000.0
    6  1991 tadult tadultStk     84009.2
    7  1991 tadult    recStk   1635970.0
    8  1991 tadult  quantity     84009.2
    9  1991  teggs   teggStk 419648000.0
    10 1991  teggs tadultStk     84009.2

为了简洁起见，您也可以像这样 Package 函数：

twostage.pivot_longer <- function(x, aggregated_names, aggregated_values,
                                      skipped_columns){
      # First aggregation
      y <- tidyr::pivot_longer(x,
                               -c(skipped_columns[[1]]),
                          names_to= aggregated_names[1],
                          values_to = aggregated_values)
      # Second aggregation
      y <- tidyr::pivot_longer(y,
                               -c(skipped_columns[[2]], aggregated_names[1]),
                               names_to= aggregated_names[2],
                               values_to = aggregated_values)
      return(y)
    }

然后像这样使用它：

annual21atest_fun <- twostage.pivot_longer(x = pivot_test,
                                           aggregated_names = c('stage', 'stock'),
                                           aggregated_values = 'quantity',
                                           skipped_columns = list(
                                             c('year', 'recStk', 'teggStk', 'tadultStk'),
                                             'year'
                                             ))