如何在 dataframe 中添加新列，如果公式在 R 編程

Question

這是我的代碼：第一列是“案例 ID”，第二列是“消息”。 我想在數據框中添加一個新列，在“消息”列中添加“歡迎”計數。

Case_id<-c("#1","#1","#1","#1","#1","#2","#2","#2","#2","#2")
message<-c("welcome to dell","welcome to dell","refresh your screen","connect to agent","Thanks good day","welcome to dell","select from default","refresh your screen","connect to agent","Thanks good day") 

df <- data.frame(Case_id, message)
    
sum(df$Case_id=="#1" & grepl("welcome*",df$message))

預期 output：

    Case id   message             Welcome message repeated
    #1        "welcome to..."         2
    #1        "welcome to..."         2
    #1        ""refresh your...."     2
    #1        ""connect to agent"     2
    #1        "Thanks good day"       2
    #2        "Welcome to..."         1
    #2         ......                 1
    #2
    #2
    #2

Answer 1

對於每個組（即Case_id ），我們可以計算組中出現“歡迎”的次數。

library(tidyverse)

df %>%
  group_by(Case_id) %>%
  mutate("Welcome message repeated" = sum(str_detect(message, "welcome*")))

或者在基礎 R 中：

transform(df, "Welcome message repeated" = ave(+(grepl("welcome*", message)), Case_id, FUN = sum))

Output

   Case_id message             `Welcome message repeated`
   <chr>   <chr>                                    <int>
 1 #1      welcome to dell                              2
 2 #1      welcome to dell                              2
 3 #1      refresh your screen                          2
 4 #1      connect to agent                             2
 5 #1      Thanks good day                              2
 6 #2      welcome to dell                              1
 7 #2      select from default                          1
 8 #2      refresh your screen                          1
 9 #2      connect to agent                             1
10 #2      Thanks good day                              1

另一種選擇是您可以使用summarise來獲取每個組的匯總計數。

df %>%
  group_by(Case_id) %>%
  summarise("Welcome message repeated" = sum(str_detect(message, "welcome*")))

#  Case_id `Welcome message repeated`
#  <chr>                        <int>
#1 #1                               2
#2 #2                               1