如何通過條件（無循環）替換data.table中的元素？

Question

我有3 data.table與在列的其餘部分第一列和數字平日。 在第一DT(DT1)，它有7行和每一列上的數字數據的n列（n>2）。其餘DT (DT2, DT3)有7行2列有一個數字列。

我想用DT3中的元素代替DT2中每個列中的元素（除了Weekdays列），其中元素在DT3中具有相同的工作日。

x = c(8.38877450980392, 7.94021071115013, 7.95032679738562, 7.44576124567474, 
8.83645276292335) 
y = c(83.8877450980392, 79.4021071115013, 79.5032679738562, 74.4576124567474, 
88.3645276292335) 
DT1 = data.table(WeekDay = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"), a = abs(rnorm(7)*100), b = abs(rnorm(7)*100), c = abs(rnorm(7)*100), d = abs(rnorm(7)*100)) 
DT2 = data.table(WeekDay = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"), criteria = x) 
DT3 = data.table(WeekDay = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"), Replace_Value = y) 

有沒有辦法做到沒有循環？ 請僅使用base R？ 謝謝。

Answer 1

這應做到：

setkey(DT1, WeekDay) 
setkey(DT2, WeekDay) 
mx.data <- as.matrix(DT1[, letters[1:4], with=F]) 
(DT1[DT2][DT3][, letters[1:4]:=as.data.table(ifelse(mx.data < criteria, Replace_Value, mx.data))]) 

首先，我合併所有的表作爲暗影建議。然後，使用I與數據在DT1創建的矩陣I使用ifelsecriteria反對，這應爲每列回收criteria。然後，我首先把它變成一個data.table到列a-d與:=更換運營商分配ifelse的結果。

影子指出，我不得不改變你的標準，所以他們有天權數。另外，我不得不讓你的標準有點大，因爲他們從來沒有（在DT1總是大於crtieria值）滿足。這是我對你的數據做的：

x = c(8.38877450980392, 7.94021071115013, 7.95032679738562, 7.44576124567474, 
     8.83645276292335, 50, 25) 
y = c(83.8877450980392, 79.4021071115013, 79.5032679738562, 74.4576124567474, 
     88.3645276292335, 50, 25) 
set.seed(1) 
DT1 = data.table(WeekDay = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"), a = abs(rnorm(7)*100), b = abs(rnorm(7)*100), c = abs(rnorm(7)*100), d = abs(rnorm(7)*100)) 
DT2 = data.table(WeekDay = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"), criteria = x * 7) 
DT3 = data.table(WeekDay = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"), Replace_Value = y)