使用依赖于 data.table 行类型的函数计算大小

Question

I want to compute a magnitude that depends on the type of a data.table row:

library (data.table)

n1 <- c(1,2,3)
n2 <- c(10,20,30,40)
num <- list(n1,n2)
type <- c("price", "volume") 
dt1 <- data.table (num, type)
dt1 [, cons.fun := "sum"]
dt1[type=="price", cons.fun := "mean"]

> dt1
           num   type cons.fun
1:       1,2,3  price     mean
2: 10,20,30,40 volume      sum

Thus, I need to create a new column magnitude based on the cons.fun value. I tried:

dt1 [, magnitude:=match.fun(cons.fun)(num)] , and got the error:

Error in match.fun(cons.fun) : 
  'c("mean", "sum")' is not a function, character or symbol

Instead of the expected:

> dt1
           num   type cons.fun magnitude
1:       1,2,3  price     mean         2
2: 10,20,30,40 volume      sum       100

How can I get it right? Thank you!

Answer 1

Here is one approach. Loop through the rows, get the value of the string and apply the function on the 'num' column

dt1[, magnitude := get(cons.fun)(num[[1]]), 1:nrow(dt1)]
dt1
#          num   type cons.fun magnitude
#1:       1,2,3  price     mean         2
#2: 10,20,30,40 volume      sum       100

---

Or as the OP was using match.fun , 'cons.fun' can be looped with mapply to extract the function object and to apply the function on the corresponding list column 'num'

dt1[, magnitude := mapply(function(x, y) match.fun(x)(y), cons.fun, num)]