Merge pull request apache#192 from cafreeman/sparkr-sql

define generic for 'first' in RDD API

pkg/NAMESPACE

@@ -17,6 +17,7 @@ exportMethods(
               "distinct",
               "Filter",
               "filterRDD",
+              "first",
               "flatMap",
               "flatMapValues",
               "fold",
@@ -87,7 +88,6 @@ exportClasses("DataFrame")
 exportMethods("columns",
               "distinct",
               "dtypes",
-              "first",
               "head",
               "limit",
               "names",

pkg/R/DataFrame.R

@@ -492,8 +492,6 @@ setMethod("head",
 #' first(df)
 #' }
 
-setGeneric("first", function(x) {standardGeneric("first") })
-
 setMethod("first",
           signature(x = "DataFrame"),
           function(x) {

pkg/R/RDD.R

@@ -841,6 +841,26 @@ setMethod("take",
             resList
           })
 
+#' First
+#'
+#' Return the first element of an RDD
+#'
+#' @rdname first
+#' @export
+#' @examples
+#'\dontrun{
+#' sc <- sparkR.init()
+#' rdd <- parallelize(sc, 1:10)
+#' first(rdd)
+#' }
+setGeneric("first", function(x) { standardGeneric("first") })
+
+setMethod("first",
+          signature(x = "RDD"),
+          function(x) {
+            take(x, 1)[[1]]
+          })
+
 #' Removes the duplicates from RDD.
 #'
 #' This function returns a new RDD containing the distinct elements in the

pkg/inst/tests/test_rdd.R

@@ -15,6 +15,12 @@ test_that("get number of partitions in RDD", {
   expect_equal(numPartitions(intRdd), 2)
 })
 
+test_that("first on RDD", {
+  expect_true(first(rdd) == 1)
+  newrdd <- lapply(rdd, function(x) x + 1)
+  expect_true(first(newrdd) == 2)  
+})
+
 test_that("count and length on RDD", {
    expect_equal(count(rdd), 10)
    expect_equal(length(rdd), 10)