#RxHiveData only works with spark compute context
computeContext <- RxSpark(consoleOutput = TRUE, persistentRun = TRUE)
rxSetComputeContext(computeContext)

airColInfo <- list(
  arrdelay =  list(type = "integer"),
  #crsdeptime = list(type = "numeric"),
  dayofweek = list(
    type = "factor",
    levels = c(
      "Monday",
      "Tuesday",
      "Wednesday",
      "Thursday",
      "Friday",
      "Saturday",
      "Sunday"
    )
  )
)

hive_data <- RxHiveData(
  query = "select * from AirlineDemoSmallHive",
  colInfo = airColInfo
)

myData <- rxDataStep(inData = hive_data, rowSelection = arrdelay > 240 & arrdelay <= 300, varsToKeep = c("arrdelay", "dayofweek"))
head(myData)