Merge pull request #15 from shivaram/sparkr-groupby-retrain

Revert workaround in SparkR to retain grouped cols
apache · May 8, 2015 · 5f923c0 · 5f923c0
2 parents b8b87e1 + c1de670
commit 5f923c0
Show file tree

Hide file tree

Showing 2 changed files with 1 addition and 14 deletions.
diff --git a/R/pkg/R/group.R b/R/pkg/R/group.R
@@ -103,9 +103,7 @@ setMethod("agg",
                 }
               }
               jcols <- lapply(cols, function(c) { c@jc })
-              # the GroupedData.agg(col, cols*) API does not contain grouping Column
-              sdf <- callJStatic("org.apache.spark.sql.api.r.SQLUtils", "aggWithGrouping",
-                                 x@sgd, listToSeq(jcols))
+              sdf <- callJMethod(x@sgd, "agg", jcols[[1]], listToSeq(jcols[-1]))
             } else {
               stop("agg can only support Column or character")
             }

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/api/r/SQLUtils.scala b/sql/core/src/main/scala/org/apache/spark/sql/api/r/SQLUtils.scala
@@ -72,17 +72,6 @@ private[r] object SQLUtils {
     sqlContext.createDataFrame(rowRDD, schema)
   }
 
-  // A helper to include grouping columns in Agg()
-  def aggWithGrouping(gd: GroupedData, exprs: Column*): DataFrame = {
-    val aggExprs = exprs.map { col =>
-      col.expr match {
-        case expr: NamedExpression => expr
-        case expr: Expression => Alias(expr, expr.simpleString)()
-      }
-    }
-    gd.toDF(aggExprs)
-  }
-
   def dfToRowRDD(df: DataFrame): JavaRDD[Array[Byte]] = {
     df.map(r => rowToRBytes(r))
   }