rxExecByKeys.R

cc <- rxSparkConnect(reset = TRUE)
hdfsFileSystem <- RxHdfsFileSystem()
textData <- RxTextData(file = "/share/AirlineDemoSmall/AirlineDemoSmall.csv", missingValueString = "M", stringsAsFactors = TRUE, fileSystem = hdfsFileSystem)
.Summary <- function(keys, data)
{
    df <- rxImport(data)
    nrow(df)
}

# single key
result <- rxExecBy(inData = textData, keys = c("DayOfWeek"), func = .Summary)

str(result)
#List of 7
# $ :List of 3
#  ..$ keys  :List of 1
#  .. ..$ : Factor w/ 7 levels "Monday","Sunday",..: 4
#  ..$ result: int 78875
#  ..$ status:List of 3
#  .. ..$ : chr "OK"
#  .. ..$ : NULL
#  .. ..$ : NULL
# $ :List of 3
#  ..$ keys  :List of 1
#  .. ..$ : Factor w/ 7 levels "Monday","Sunday",..: 6
#  ..$ result: int 82987
#  ..$ status:List of 3
#  .. ..$ : chr "OK"
#  .. ..$ : NULL
#  .. ..$ : NULL
# $ :List of 3
#  ..$ keys  :List of 1
#  .. ..$ : Factor w/ 7 levels "Monday","Sunday",..: 5
#  ..$ result: int 81304
#  ..$ status:List of 3
#  .. ..$ : chr "OK"
#  .. ..$ : NULL
#  .. ..$ : NULL
# $ :List of 3
#  ..$ keys  :List of 1
#  .. ..$ : Factor w/ 7 levels "Monday","Sunday",..: 7
#  ..$ result: int 86159
#  ..$ status:List of 3
#  .. ..$ : chr "OK"
#  .. ..$ : NULL
#  .. ..$ : NULL
# $ :List of 3
#  ..$ keys  :List of 1
#  .. ..$ : Factor w/ 7 levels "Monday","Sunday",..: 1
#  ..$ result: int 97975
#  ..$ status:List of 3
#  .. ..$ : chr "OK"
#  .. ..$ : NULL
#  .. ..$ : NULL
# $ :List of 3
#  ..$ keys  :List of 1
#  .. ..$ : Factor w/ 7 levels "Monday","Sunday",..: 3
#  ..$ result: int 77725
#  ..$ status:List of 3
#  .. ..$ : chr "OK"
#  .. ..$ : NULL
#  .. ..$ : NULL
# $ :List of 3
#  ..$ keys  :List of 1
#  .. ..$ : Factor w/ 7 levels "Monday","Sunday",..: 2
#  ..$ result: int 94975
#  ..$ status:List of 3
#  .. ..$ : chr "OK"
#  .. ..$ : NULL
#  .. ..$ : NULL

# multi keys
result <- rxExecBy(inData = textData, keys = c("DayOfWeek", "ArrDelay"), func = .Summary)

length(result)
#[1] 3233

str(result[[1]])
#List of 3
# $ keys  :List of 2
#  ..$ : Factor w/ 7 levels "Monday","Sunday",..: 4
#  ..$ : int 388
# $ result: int 2
# $ status:List of 3
#  ..$ : chr "OK"
#  ..$ : NULL
#  ..$ : NULL

rxSparkDisconnect(cc)