Aggregating a dataframe by a group of columns passed as arguments

Question

I have a dataframe which contains feature set for a person at a different times of the day. I want to aggregate it for different durations of time ie 1 hr, 2 hr, and so on. I have written this function as follows.

 library(data.table)

 getAggregate <- function(comb,limb) {
      comb[Limb == as.character(limb),
      c(fb(BC),
      structure(lapply(.SD,mean),names=meanFeatures),
      structure(lapply(.SD,ivar),names=varFeatures)
      )
     ,by=.(Person,hour(Date),yday(Date),wday(Date))
     ,.SDcols = sigfeats]
 }

I just want to pass "Person,hour(Date),yday(Date),wday(Date)" as arguments to the function just like "limb" so that it can aggregate using these four values (or any other values incase the arguments are different).

Answer 1

Maybe this would work like this...

library(lubridate)
library(data.table)

getAggregate <- function(comb, limb, byList) {
    comb$Hour <- hour(comb$Date)
    comb$YDay <- yday(comb$Date)
    comb$WDay <- wday(comb$Date)
    comb$Minute <- minute(Date)

    comb[Limb == as.character(limb),
         c(
             fb(BC),
             structure(lapply(.SD, mean), names = meanFeatures),
             structure(lapply(.SD, ivar), names = varFeatures)
         )
         , by = byList
         , .SDcols = sigfeats]
}

...and you would call the function like so...

getAggregate(comb, limb, byList = c("Person", "Hour", "YDay", "WDay", "Minute"))