Question

我有一个像这样的数据框

library(plyr)
library(dplyr)

ID <- c("ID001","ID002","ID003","ID004","ID005",
        "ID006","ID007","ID008","ID009","ID010")
Fail <- c(3,7,2,3,9,7,3,2,3,9)
Pass <- c(0,0,4,26,1,1,3,0,9,9)
df <- data.frame(ID,Fail,Pass)

我添加了另一列来计算失败百分比

df$Fail_Percentage <- (df$Fail/(df$Fail+df$Pass))*100

现在，我订购此数据框并创建变量“Rank”

library(data.table)
df <- df%>%
  arrange(-Fail_Percentage) %>%
  mutate(Rank = rleid(Fail_Percentage))
df

我得到了这个输出

      ID Fail Pass Fail_Percentage Rank
   ID001    3    0       100.00000    1
   ID002    7    0       100.00000    1
   ID008    2    0       100.00000    1
   ID005    9    1        90.00000    2
   ID006    7    1        87.50000    3
   ID007    3    3        50.00000    4
   ID010    9    9        50.00000    4
   ID003    2    4        33.33333    5
   ID009    3    9        25.00000    6
   ID004    3   26        10.34483    7

这里的问题是我以这种方式创建重复的排名。我希望通过优先选择“失败”来排名。

例如：ID001，ID002，ID008都有排名1，但我想通过优先顺序排列最高失败率排名。因此，ID002将具有等级1，ID001将具有等级2并且ID008将具有等级3.我想以这种方式这样做并且同样地对其他条目进行排名。

所需的输出将是

          ID Fail Pass Fail_Percentage Rank
       ID002    7    0       100.00000    1
       ID001    3    0       100.00000    2
       ID008    2    0       100.00000    3
       ID005    9    1        90.00000    4
       ID006    7    1        87.50000    5
       ID010    9    9        50.00000    6
       ID007    3    3        50.00000    7
       ID003    2    4        33.33333    8
       ID009    3    9        25.00000    9
       ID004    3   26        10.34483    10

我们怎样才能更好地做到这一点？有人能帮我指出正确的方向吗？

Answer 1

ID <- c("ID001","ID002","ID003","ID004","ID005",
        "ID006","ID007","ID008","ID009","ID010")
Fail <- c(3,7,2,3,9,7,3,2,3,9)
Pass <- c(0,0,4,26,1,1,3,0,9,9)
df <- data.frame(ID,Fail,Pass)
df$Fail_Percentage <- (df$Fail/(df$Fail+df$Pass))*100

仅使用data.table

df <- setDT(df)[order(-Fail_Percentage, -Fail)][, Rank := 1:.N]

Answer 2

您可以使用第二个参数来安排预定的订单：

library(dplyr)

df = structure(list(ID = structure(1:10, .Label = c("ID001", "ID002", 
"ID003", "ID004", "ID005", "ID006", "ID007", "ID008", "ID009", 
"ID010"), class = "factor"), Fail = c(3, 7, 2, 3, 9, 7, 3, 2, 
3, 9), Pass = c(0, 0, 4, 26, 1, 1, 3, 0, 9, 9)), .Names = c("ID", 
"Fail", "Pass"), row.names = c(NA, -10L), class = "data.frame") 

df = df %>%
    mutate(Fail_Percentage = Fail / (Fail + Pass) * 100) %>%
    arrange(-Fail_Percentage, -Fail) %>%
    mutate(Rank = order(-Fail_Percentage))

> df

ID Fail Pass Fail_Percentage Rank
1  ID002    7    0       100.00000    1
2  ID001    3    0       100.00000    2
3  ID008    2    0       100.00000    3
4  ID005    9    1        90.00000    4
5  ID006    7    1        87.50000    5
6  ID010    9    9        50.00000    6
7  ID007    3    3        50.00000    7
8  ID003    2    4        33.33333    8
9  ID009    3    9        25.00000    9
10 ID004    3   26        10.34483   10

如何使用失败百分比对ID进行排名

2 个答案: