在 R 中使用 pivot_wider() 使用 values_fn 执行聚合（例如求和、平均值）时出错

Question

我有一个具有以下格式的数据集：

> library(tidyverse)
> library(tibble)
> 
> 
> data<-data.frame(ID=c(1,1,2,2,3,3,3,3,4,4),
+          Radius=c(5,5,5,5,10,10,15,15,10,10),
+          neighb_ID=c(1,11,2,12,3,4,7,8,3,4),
+          var_neighb=c(50,20,30,40,15,100,70,60,15,100))
> data
   ID1 Radius neighb_ID var_neighb
1    1      5         1         50
2    1      5        11         20
3    2      5         2         30
4    2      5        12         40
5    3     10         3         15
6    3     10         4        100
7    3     15         7         70
8    3     15         8         60
9    4     10         3         15
10   4     10         4        100
> 

现在我想 pivot 这个数据，以便为每个ID按Radius聚合var_neighb 。 例如，对于sum和mean ，我希望实现下表：

  ID1 Svar_neighb_Radius_5 Svar_neighb_Radius_10 Svar_neighb_Radius_15
1   1                   20                     0                     0
2   2                   40                     0                     0
3   3                    0                   100                   130
4   4                    0                    15                     0
  Mvar_neighb_Radius_5 Mvar_neighb_Radius_10 Mvar_neighb_Radius_15
1                   20                     0                     0
2                   40                     0                     0
3                    0                   100                    65
4                    0                    15                     0
> 

我尝试使用以下代码执行此操作：

> agdata<-data %>%
+    pivot_wider(
+     names_from = Radius, 
+     values_from = var_neighb,
+     values_fn = sum,
+     values_fill = 0
+     )

我只收到以下错误：

Error in values_fn[[value]] : object of type 'builtin' is not subsettable

此外，即使我取出values_fn = sum, ，我也会收到以下错误： Error in values_fill[[value]]: subscript out of bounds 。

有人可以帮助我解决这些问题以实现我的目标吗？

编辑：对不起，我忽略了 output 表的一个重要要求：聚合应该是sum和mean并且不应该包括var_neighb的值，其中neighb_ID等于ID 。 output 表data_out需要按sum和mean进行聚合。 所以我更新了data 。

Answer 1

values_fn和values_fill应该命名为列表：

library(tidyverse)

data <- data.frame(
  ID=c(1,1,2,2,3,3,3,4,4),
  Radius=c(5,5,5,5,10,10,15,10,10),
  neighb_ID=c(1,11,2,12,3,4,7,3,4),
  var_neighb=c(50,20,30,40,15,100,70,15,100)
)

data %>%
  select(-neighb_ID) %>%
  pivot_wider(
     names_from = Radius, 
     values_from = var_neighb,
     values_fn = list(var_neighb = sum),
     values_fill = list(var_neighb = 0),
     names_prefix = "var_neighb_Radius_"
   )

# # A tibble: 4 x 4
#       ID                var_neighb_Radius_5   var_neighb_Radius_10 var_neighb_Radius_15
#       <dbl>              <dbl>                <dbl>                <dbl>
# 1     1                  70                    0                    0
# 2     2                  70                    0                    0
# 3     3                   0                  115                   70
# 4     4                   0                  115                    0

更新要删除ID == neighb_ID的值，只需使用过滤器：

data %>%
  filter(ID != neighb_ID) %>%
  select(-neighb_ID) %>%
  pivot_wider(
     names_from = Radius, 
     values_from = var_neighb,
     values_fn = list(var_neighb = sum),
     values_fill = list(var_neighb = 0),
     names_prefix = "var_neighb_Radius_"
   )

不太确定您对“按mean和按sum进行聚合”的理解 - 您不能在单个列中进行两个不同的聚合，但您可以制作两个枢轴并将它们连接在一起：

library(dplyr)

inner_join(
  data %>%
    filter(ID != neighb_ID) %>%
    select(-neighb_ID) %>%
    pivot_wider(
      names_from = Radius, 
      values_from = var_neighb,
      values_fn = list(var_neighb = sum),
      values_fill = list(var_neighb = 0),
      names_prefix = "var_neighb_Radius_sum_"
    ),
  data %>%
    filter(ID != neighb_ID) %>%
    select(-neighb_ID) %>%
    pivot_wider(
      names_from = Radius, 
      values_from = var_neighb,
      values_fn = list(var_neighb = mean),
      values_fill = list(var_neighb = 0),
      names_prefix = "var_neighb_Radius_mean_"
    ),
  by = "ID"
)