Support More Aggregate Methods (#265)

src/csharp/Microsoft.Spark.E2ETest/IpcTests/Sql/DataFrameTests.cs

@@ -404,9 +404,25 @@ public void TestSignaturesV2_3_X()
             Assert.IsType<RelationalGroupedDataset>(_df.GroupBy(_df["age"]));
             Assert.IsType<RelationalGroupedDataset>(_df.GroupBy(_df["age"], _df["name"]));
 
-            Assert.IsType<DataFrame>(_df.GroupBy("name").Sum("age"));
-            Assert.IsType<DataFrame>(
-                _df.WithColumn("tempAge", _df["age"]).GroupBy("name").Sum("age", "tempAge"));
+            {
+                RelationalGroupedDataset df = 
+                    _df.WithColumn("tempAge", _df["age"]).GroupBy("name");
+
+                Assert.IsType<DataFrame>(df.Mean("age"));
+                Assert.IsType<DataFrame>(df.Mean("age", "tempAge"));
+
+                Assert.IsType<DataFrame>(df.Max("age"));
+                Assert.IsType<DataFrame>(df.Max("age", "tempAge"));
+
+                Assert.IsType<DataFrame>(df.Avg("age"));
+                Assert.IsType<DataFrame>(df.Avg("age", "tempAge"));
+
+                Assert.IsType<DataFrame>(df.Min("age"));
+                Assert.IsType<DataFrame>(df.Min("age", "tempAge"));
+
+                Assert.IsType<DataFrame>(df.Sum("age"));
+                Assert.IsType<DataFrame>(df.Sum("age", "tempAge"));
+            }
 
             Assert.IsType<RelationalGroupedDataset>(_df.Rollup("age"));
             Assert.IsType<RelationalGroupedDataset>(_df.Rollup("age", "name"));

src/csharp/Microsoft.Spark/Sql/RelationalGroupedDataset.cs

@@ -44,9 +44,42 @@ public DataFrame Agg(Column expr, params Column[] exprs) =>
         public DataFrame Count() =>
             new DataFrame((JvmObjectReference)_jvmObject.Invoke("count"));
 
+        /// <summary>
+        /// Compute the mean value for each numeric columns for each group.
+        /// </summary>
+        /// <param name="colNames">Name of columns to compute mean on</param>
+        /// <returns>New DataFrame object with mean applied</returns>
+        public DataFrame Mean(params string[] colNames) =>
+            new DataFrame((JvmObjectReference)_jvmObject.Invoke("mean", (object)colNames));
+
+        /// <summary>
+        /// Compute the max value for each numeric columns for each group.
+        /// </summary>
+        /// <param name="colNames">Name of columns to compute max on</param>
+        /// <returns>New DataFrame object with max applied</returns>
+        public DataFrame Max(params string[] colNames) =>
+            new DataFrame((JvmObjectReference)_jvmObject.Invoke("max", (object)colNames));
+
+        /// <summary>
+        /// Compute the average value for each numeric columns for each group.
+        /// </summary>
+        /// <param name="colNames">Name of columns to compute average on</param>
+        /// <returns>New DataFrame object with average applied</returns>
+        public DataFrame Avg(params string[] colNames) =>
+            new DataFrame((JvmObjectReference)_jvmObject.Invoke("avg", (object)colNames));
+
+        /// <summary>
+        /// Compute the min value for each numeric columns for each group.
+        /// </summary>
+        /// <param name="colNames">Name of columns to compute min on</param>
+        /// <returns>New DataFrame object with min applied</returns>
+        public DataFrame Min(params string[] colNames) =>
+            new DataFrame((JvmObjectReference)_jvmObject.Invoke("min", (object)colNames));
+
         /// <summary>
         /// Compute the sum for each numeric columns for each group.
         /// </summary>
+        /// <param name="colNames">Name of columns to compute sum on</param>
         /// <returns>New DataFrame object with sum applied</returns>
         public DataFrame Sum(params string[] colNames) =>
             new DataFrame((JvmObjectReference)_jvmObject.Invoke("sum", (object)colNames));