这个是 moving average 的内容吧

install.packages('TTR')
library(dplyr)
library(TTR)

df <- data.frame(index = 1:20, name = c("A", "B", "C", "A", "A", "C", "B", "C", "A", "B", "C", "A", "A", "A", "C", "B", "C", "A", "B", "C"), amount = c(10,14,3,4,15,9,12,6,7,8,10,13,12,6,8,8,9,3,14,10))

all_data=data.frame()
for (a in unique(df$name)){
data=df%>%filter(name==a)%>%mutate(last3avg=runMean(amount, n = 3, cumulative = FALSE))  
all_data=rbind(data,all_data)
}

all_data=all_data%>%arrange(index)
all_data

感謝 rainningpoet 兄的回答, 效果是我想要的, 謝謝.

另外, 亦多謝jinkelazzz兄在EXCEL 的見解,
我提問沒有說清楚是用 R, 真不好意思, 但十分多謝熱心的前輩

rainningpoet 兄,

df <- data.frame(index = 1:20, name = c("A", "B", "C", "A", "A", "C", "B", "C", "A", "B", "C", "A", "A", "A", "C", "B", "C", "A", "B", "D"), amount = c(10,14,3,4,15,9,12,6,7,8,10,13,12,6,8,8,9,3,14,10))

如果我name第20項, 改為"D", 你的公式就沒法远作了, 請問有其他解決方法嗎 ?

因为D不够3位。

library(dplyr)
library(TTR)

df <- data.frame(index = 1:20, name = c("A", "B", "C", "A", "A", "C", "B", "C", "A", "B", "C", "A", "A", "A", "C", "B", "C", "A", "B", "D"),
                 amount = c(10,14,3,4,15,9,12,6,7,8,10,13,12,6,8,8,9,3,14,10))


all_data=data.frame()
for (a in unique(df$name)){
  if (nrow(df%>%filter(name==a))<3){
  data=df%>%filter(name==a)%>%mutate(last3avg=NA)
  }else{
  data=df%>%filter(name==a)%>%mutate(last3avg=runMean(amount, n = 3, cumulative = FALSE))
  }
  all_data=rbind(data,all_data)
}

all_data=all_data%>%arrange(index)
all_data