diff options
author | Forest Fang <forest.fang@outlook.com> | 2015-12-29 12:45:24 +0530 |
---|---|---|
committer | Shivaram Venkataraman <shivaram@cs.berkeley.edu> | 2015-12-29 12:45:24 +0530 |
commit | d80cc90b5545cff82cd9b340f12d01eafc9ca524 (patch) | |
tree | 2f4894084bc072a93323a6f57ececd8311d68c8a /R/pkg | |
parent | 73862a1eb9744c3c32458c9c6f6431c23783786a (diff) | |
download | spark-d80cc90b5545cff82cd9b340f12d01eafc9ca524.tar.gz spark-d80cc90b5545cff82cd9b340f12d01eafc9ca524.tar.bz2 spark-d80cc90b5545cff82cd9b340f12d01eafc9ca524.zip |
[SPARK-12526][SPARKR] ifelse`, `when`, `otherwise` unable to take Column as value
`ifelse`, `when`, `otherwise` is unable to take `Column` typed S4 object as values.
For example:
```r
ifelse(lit(1) == lit(1), lit(2), lit(3))
ifelse(df$mpg > 0, df$mpg, 0)
```
will both fail with
```r
attempt to replicate an object of type 'environment'
```
The PR replaces `ifelse` calls with `if ... else ...` inside the function implementations to avoid attempt to vectorize(i.e. `rep()`). It remains to be discussed whether we should instead support vectorization in these functions for consistency because `ifelse` in base R is vectorized but I cannot foresee any scenarios these functions will want to be vectorized in SparkR.
For reference, added test cases which trigger failures:
```r
. Error: when(), otherwise() and ifelse() with column on a DataFrame ----------
error in evaluating the argument 'x' in selecting a method for function 'collect':
error in evaluating the argument 'col' in selecting a method for function 'select':
attempt to replicate an object of type 'environment'
Calls: when -> when -> ifelse -> ifelse
1: withCallingHandlers(eval(code, new_test_environment), error = capture_calls, message = function(c) invokeRestart("muffleMessage"))
2: eval(code, new_test_environment)
3: eval(expr, envir, enclos)
4: expect_equal(collect(select(df, when(df$a > 1 & df$b > 2, lit(1))))[, 1], c(NA, 1)) at test_sparkSQL.R:1126
5: expect_that(object, equals(expected, label = expected.label, ...), info = info, label = label)
6: condition(object)
7: compare(actual, expected, ...)
8: collect(select(df, when(df$a > 1 & df$b > 2, lit(1))))
Error: Test failures
Execution halted
```
Author: Forest Fang <forest.fang@outlook.com>
Closes #10481 from saurfang/spark-12526.
Diffstat (limited to 'R/pkg')
-rw-r--r-- | R/pkg/R/column.R | 4 | ||||
-rw-r--r-- | R/pkg/R/functions.R | 13 | ||||
-rw-r--r-- | R/pkg/inst/tests/testthat/test_sparkSQL.R | 8 |
3 files changed, 18 insertions, 7 deletions
diff --git a/R/pkg/R/column.R b/R/pkg/R/column.R index 7bb8ef2595..356bcee3cf 100644 --- a/R/pkg/R/column.R +++ b/R/pkg/R/column.R @@ -215,7 +215,7 @@ setMethod("%in%", #' otherwise #' -#' If values in the specified column are null, returns the value. +#' If values in the specified column are null, returns the value. #' Can be used in conjunction with `when` to specify a default value for expressions. #' #' @rdname otherwise @@ -225,7 +225,7 @@ setMethod("%in%", setMethod("otherwise", signature(x = "Column", value = "ANY"), function(x, value) { - value <- ifelse(class(value) == "Column", value@jc, value) + value <- if (class(value) == "Column") { value@jc } else { value } jc <- callJMethod(x@jc, "otherwise", value) column(jc) }) diff --git a/R/pkg/R/functions.R b/R/pkg/R/functions.R index 09e4e04335..df36bc869a 100644 --- a/R/pkg/R/functions.R +++ b/R/pkg/R/functions.R @@ -37,7 +37,7 @@ setMethod("lit", signature("ANY"), function(x) { jc <- callJStatic("org.apache.spark.sql.functions", "lit", - ifelse(class(x) == "Column", x@jc, x)) + if (class(x) == "Column") { x@jc } else { x }) column(jc) }) @@ -2262,7 +2262,7 @@ setMethod("unix_timestamp", signature(x = "Column", format = "character"), setMethod("when", signature(condition = "Column", value = "ANY"), function(condition, value) { condition <- condition@jc - value <- ifelse(class(value) == "Column", value@jc, value) + value <- if (class(value) == "Column") { value@jc } else { value } jc <- callJStatic("org.apache.spark.sql.functions", "when", condition, value) column(jc) }) @@ -2277,13 +2277,16 @@ setMethod("when", signature(condition = "Column", value = "ANY"), #' @name ifelse #' @seealso \link{when} #' @export -#' @examples \dontrun{ifelse(df$a > 1 & df$b > 2, 0, 1)} +#' @examples \dontrun{ +#' ifelse(df$a > 1 & df$b > 2, 0, 1) +#' ifelse(df$a > 1, df$a, 1) +#' } setMethod("ifelse", signature(test = "Column", yes = "ANY", no = "ANY"), function(test, yes, no) { test <- test@jc - yes <- ifelse(class(yes) == "Column", yes@jc, yes) - no <- ifelse(class(no) == "Column", no@jc, no) + yes <- if (class(yes) == "Column") { yes@jc } else { yes } + no <- if (class(no) == "Column") { no@jc } else { no } jc <- callJMethod(callJStatic("org.apache.spark.sql.functions", "when", test, yes), diff --git a/R/pkg/inst/tests/testthat/test_sparkSQL.R b/R/pkg/inst/tests/testthat/test_sparkSQL.R index 135c7576e5..c2b6adbe3a 100644 --- a/R/pkg/inst/tests/testthat/test_sparkSQL.R +++ b/R/pkg/inst/tests/testthat/test_sparkSQL.R @@ -1120,6 +1120,14 @@ test_that("when(), otherwise() and ifelse() on a DataFrame", { expect_equal(collect(select(df, ifelse(df$a > 1 & df$b > 2, 0, 1)))[, 1], c(1, 0)) }) +test_that("when(), otherwise() and ifelse() with column on a DataFrame", { + l <- list(list(a = 1, b = 2), list(a = 3, b = 4)) + df <- createDataFrame(sqlContext, l) + expect_equal(collect(select(df, when(df$a > 1 & df$b > 2, lit(1))))[, 1], c(NA, 1)) + expect_equal(collect(select(df, otherwise(when(df$a > 1, lit(1)), lit(0))))[, 1], c(0, 1)) + expect_equal(collect(select(df, ifelse(df$a > 1 & df$b > 2, lit(0), lit(1))))[, 1], c(1, 0)) +}) + test_that("group by, agg functions", { df <- read.json(sqlContext, jsonPath) df1 <- agg(df, name = "max", age = "sum") |