Skip to content

[SPARK-21727][R] Allow multi-element atomic vector as column type in SparkR DataFrame #20352

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 3 commits into from
Closed

[SPARK-21727][R] Allow multi-element atomic vector as column type in SparkR DataFrame #20352

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6bdf687
Check if an atomic R type is actually a vector of length > 1, and tre…
neilalex Jan 5, 2018
f8ae698
Use is.atomic(object) to check type
neilalex Jan 22, 2018
01fc9e1
tests for SparkR data.frame with multi-element atomic vector
neilalex Jan 23, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 6 additions & 5 deletions R/pkg/R/serialize.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,14 +30,17 @@
# POSIXct,POSIXlt -> Time
#
# list[T] -> Array[T], where T is one of above mentioned types
# Multi-element vector of any of the above (except raw) -> Array[T]
# environment -> Map[String, T], where T is a native type
# jobj -> Object, where jobj is an object created in the backend
# nolint end

getSerdeType <- function(object) {
type <- class(object)[[1]]
if (type != "list") {
type
if (is.atomic(object) & !is.raw(object) & length(object) > 1) {
"array"
} else if (type != "list") {
type
} else {
# Check if all elements are of same type
elemType <- unique(sapply(object, function(elem) { getSerdeType(elem) }))
Expand All @@ -50,9 +53,7 @@ getSerdeType <- function(object) {
}

writeObject <- function(con, object, writeType = TRUE) {
# NOTE: In R vectors have same type as objects. So we don't support
# passing in vectors as arrays and instead require arrays to be passed
# as lists.
# NOTE: In R vectors have same type as objects
type <- class(object)[[1]] # class of POSIXlt is c("POSIXlt", "POSIXt")
# Checking types is needed here, since 'is.na' only handles atomic vectors,
# lists and pairlists
Expand Down
47 changes: 47 additions & 0 deletions R/pkg/tests/fulltests/test_Serde.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,6 +37,53 @@ test_that("SerDe of primitive types", {
expect_equal(class(x), "character")
})

test_that("SerDe of multi-element primitive vectors inside R data.frame", {
# vector of integers embedded in R data.frame
indices <- 1L:3L
myDf <- data.frame(indices)
myDf$data <- list(rep(0L, 3L))
mySparkDf <- as.DataFrame(myDf)
myResultingDf <- collect(mySparkDf)
myDfListedData <- data.frame(indices)
myDfListedData$data <- list(as.list(rep(0L, 3L)))
expect_equal(myResultingDf, myDfListedData)
expect_equal(class(myResultingDf[["data"]][[1]]), "list")
expect_equal(class(myResultingDf[["data"]][[1]][[1]]), "integer")

# vector of numeric embedded in R data.frame
myDf <- data.frame(indices)
myDf$data <- list(rep(0, 3L))
mySparkDf <- as.DataFrame(myDf)
myResultingDf <- collect(mySparkDf)
myDfListedData <- data.frame(indices)
myDfListedData$data <- list(as.list(rep(0, 3L)))
expect_equal(myResultingDf, myDfListedData)
expect_equal(class(myResultingDf[["data"]][[1]]), "list")
expect_equal(class(myResultingDf[["data"]][[1]][[1]]), "numeric")

# vector of logical embedded in R data.frame
myDf <- data.frame(indices)
myDf$data <- list(rep(TRUE, 3L))
mySparkDf <- as.DataFrame(myDf)
myResultingDf <- collect(mySparkDf)
myDfListedData <- data.frame(indices)
myDfListedData$data <- list(as.list(rep(TRUE, 3L)))
expect_equal(myResultingDf, myDfListedData)
expect_equal(class(myResultingDf[["data"]][[1]]), "list")
expect_equal(class(myResultingDf[["data"]][[1]][[1]]), "logical")

# vector of character embedded in R data.frame
myDf <- data.frame(indices)
myDf$data <- list(rep("abc", 3L))
mySparkDf <- as.DataFrame(myDf)
myResultingDf <- collect(mySparkDf)
myDfListedData <- data.frame(indices)
myDfListedData$data <- list(as.list(rep("abc", 3L)))
expect_equal(myResultingDf, myDfListedData)
expect_equal(class(myResultingDf[["data"]][[1]]), "list")
expect_equal(class(myResultingDf[["data"]][[1]][[1]]), "character")
})

test_that("SerDe of list of primitive types", {
x <- list(1L, 2L, 3L)
y <- callJStatic("SparkRHandler", "echo", x)
Expand Down