diff options
| author | Sun Rui <rui.sun@intel.com> | 2015-08-16 00:30:02 -0700 |
|---|---|---|
| committer | Shivaram Venkataraman <shivaram@cs.berkeley.edu> | 2015-08-16 00:30:02 -0700 |
| commit | 5f9ce738fe6bab3f0caffad0df1d3876178cf469 (patch) | |
| tree | 4fa3cb7483492e30f7c1f2951b010f24b6da901c /R | |
| parent | 182f9b7a6d3a3ee7ec7de6abc24e296aa794e4e8 (diff) | |
| download | spark-5f9ce738fe6bab3f0caffad0df1d3876178cf469.tar.gz spark-5f9ce738fe6bab3f0caffad0df1d3876178cf469.tar.bz2 spark-5f9ce738fe6bab3f0caffad0df1d3876178cf469.zip | |
[SPARK-8844] [SPARKR] head/collect is broken in SparkR.
This is a WIP patch for SPARK-8844 for collecting reviews.
This bug is about reading an empty DataFrame. in readCol(),
lapply(1:numRows, function(x) {
does not take into consideration the case where numRows = 0.
Will add unit test case.
Author: Sun Rui <rui.sun@intel.com>
Closes #7419 from sun-rui/SPARK-8844.
Diffstat (limited to 'R')
| -rw-r--r-- | R/pkg/R/deserialize.R | 16 | ||||
| -rw-r--r-- | R/pkg/inst/tests/test_sparkSQL.R | 20 |
2 files changed, 30 insertions, 6 deletions
diff --git a/R/pkg/R/deserialize.R b/R/pkg/R/deserialize.R index 6d364f77be..33bf13ec9e 100644 --- a/R/pkg/R/deserialize.R +++ b/R/pkg/R/deserialize.R @@ -176,10 +176,14 @@ readRow <- function(inputCon) { # Take a single column as Array[Byte] and deserialize it into an atomic vector readCol <- function(inputCon, numRows) { - # sapply can not work with POSIXlt - do.call(c, lapply(1:numRows, function(x) { - value <- readObject(inputCon) - # Replace NULL with NA so we can coerce to vectors - if (is.null(value)) NA else value - })) + if (numRows > 0) { + # sapply can not work with POSIXlt + do.call(c, lapply(1:numRows, function(x) { + value <- readObject(inputCon) + # Replace NULL with NA so we can coerce to vectors + if (is.null(value)) NA else value + })) + } else { + vector() + } } diff --git a/R/pkg/inst/tests/test_sparkSQL.R b/R/pkg/inst/tests/test_sparkSQL.R index e6d3b21ff8..c77f633135 100644 --- a/R/pkg/inst/tests/test_sparkSQL.R +++ b/R/pkg/inst/tests/test_sparkSQL.R @@ -408,6 +408,14 @@ test_that("collect() returns a data.frame", { expect_equal(names(rdf)[1], "age") expect_equal(nrow(rdf), 3) expect_equal(ncol(rdf), 2) + + # collect() returns data correctly from a DataFrame with 0 row + df0 <- limit(df, 0) + rdf <- collect(df0) + expect_true(is.data.frame(rdf)) + expect_equal(names(rdf)[1], "age") + expect_equal(nrow(rdf), 0) + expect_equal(ncol(rdf), 2) }) test_that("limit() returns DataFrame with the correct number of rows", { @@ -492,6 +500,18 @@ test_that("head() and first() return the correct data", { testFirst <- first(df) expect_equal(nrow(testFirst), 1) + + # head() and first() return the correct data on + # a DataFrame with 0 row + df0 <- limit(df, 0) + + testHead <- head(df0) + expect_equal(nrow(testHead), 0) + expect_equal(ncol(testHead), 2) + + testFirst <- first(df0) + expect_equal(nrow(testFirst), 0) + expect_equal(ncol(testFirst), 2) }) test_that("distinct() and unique on DataFrames", { |