[GitHub] [arrow] jorisvandenbossche commented on a change in pull request #7920: ARROW-9638: [C++][Compute] Implement mode kernel

[GitBox <gi...@apache.org>]

jorisvandenbossche commented on a change in pull request #7920:
URL: https://github.com/apache/arrow/pull/7920#discussion_r469186653



##########
File path: cpp/src/arrow/compute/kernels/aggregate_test.cc
##########
@@ -594,5 +594,96 @@ TYPED_TEST(TestFloatingMinMaxKernel, DefaultOptions) {
   AssertDatumsEqual(explicit_defaults, no_options_provided);
 }
 
+//
+// Mode
+//
+
+template <typename ArrowType>
+class TestPrimitiveModeKernel : public ::testing::Test {
+  using Traits = TypeTraits<ArrowType>;
+  using c_type = typename ArrowType::c_type;
+  using ModeType = typename Traits::ScalarType;
+  using CountType = typename TypeTraits<Int64Type>::ScalarType;
+
+ public:
+  void AssertModeIs(const Datum& array, c_type expected_mode, int64_t expected_count) {
+    ASSERT_OK_AND_ASSIGN(Datum out, Mode(array));
+    const StructScalar& value = out.scalar_as<StructScalar>();
+
+    const auto& out_mode = checked_cast<const ModeType&>(*value.value[0]);
+    ASSERT_EQ(expected_mode, out_mode.value);
+
+    const auto& out_count = checked_cast<const CountType&>(*value.value[1]);
+    ASSERT_EQ(expected_count, out_count.value);
+  }
+
+  void AssertModeIs(const std::string& json, c_type expected_mode,
+                    int64_t expected_count) {
+    auto array = ArrayFromJSON(type_singleton(), json);
+    AssertModeIs(array, expected_mode, expected_count);
+  }
+
+  void AssertModeIsNull(const Datum& array) {
+    ASSERT_OK_AND_ASSIGN(Datum out, Mode(array));
+    const StructScalar& value = out.scalar_as<StructScalar>();
+
+    for (const auto& val : value.value) {
+      ASSERT_FALSE(val->is_valid);
+    }
+  }
+
+  void AssertModeIsNull(const std::string& json) {
+    auto array = ArrayFromJSON(type_singleton(), json);
+    AssertModeIsNull(array);
+  }
+
+  std::shared_ptr<DataType> type_singleton() { return Traits::type_singleton(); }
+};
+
+template <typename ArrowType>
+class TestIntegerModeKernel : public TestPrimitiveModeKernel<ArrowType> {};
+
+template <typename ArrowType>
+class TestFloatingModeKernel : public TestPrimitiveModeKernel<ArrowType> {};
+
+class TestBooleanModeKernel : public TestPrimitiveModeKernel<BooleanType> {};
+
+TEST_F(TestBooleanModeKernel, Basics) {
+  this->AssertModeIs("[false, false]", false, 2);
+  this->AssertModeIs("[false, false, true, true, true]", true, 3);
+  this->AssertModeIs("[true, false, false, true, true]", true, 3);
+  this->AssertModeIs("[false, false, true, true, true, false]", false, 3);
+
+  this->AssertModeIs("[true, null, false, false, null, true, null, null, true]", true, 3);
+  this->AssertModeIsNull("[null, null, null]");
+  this->AssertModeIsNull("[]");
+}
+
+TYPED_TEST_SUITE(TestIntegerModeKernel, IntegralArrowTypes);
+TYPED_TEST(TestIntegerModeKernel, Basics) {
+  this->AssertModeIs("[5, 1, 1, 5, 5]", 5, 3);
+  this->AssertModeIs("[5, 1, 1, 5, 5, 1]", 1, 3);
+  this->AssertModeIs("[127, 0, 127, 127, 0, 1, 0, 127]", 127, 4);
+
+  this->AssertModeIs("[null, null, 2, null, 1]", 1, 1);
+  this->AssertModeIsNull("[null, null, null]");
+  this->AssertModeIsNull("[]");
+}
+
+TYPED_TEST_SUITE(TestFloatingModeKernel, RealArrowTypes);
+TYPED_TEST(TestFloatingModeKernel, Floats) {
+  this->AssertModeIs("[5, 1, 1, 5, 5]", 5, 3);
+  this->AssertModeIs("[5, 1, 1, 5, 5, 1]", 1, 3);
+  this->AssertModeIs("[Inf, 100, Inf, 100, Inf]", INFINITY, 3);
+  this->AssertModeIs("[Inf, -Inf, Inf, -Inf]", -INFINITY, 2);
+
+  this->AssertModeIs("[null, null, 2, null, 1]", 1, 1);
+  this->AssertModeIs("[NaN, NaN, 1]", 1, 1);
+  this->AssertModeIsNull("[null, null, null]");
+  this->AssertModeIsNull("[NaN, NaN, null]");
+  this->AssertModeIsNull("[NaN, NaN, NaN]");

Review comment:
       Should this return NaN instead of Null?




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org