Removed const path for arrayAggregation

This commit is contained in:
Maksim Kita 2021-04-30 12:33:34 +03:00
parent aa71b4a6df
commit 44b966af5a
3 changed files with 53 additions and 130 deletions

View File

@ -132,131 +132,47 @@ struct ArrayAggregateImpl
template <typename Element>
static NO_SANITIZE_UNDEFINED bool executeType(const ColumnPtr & mapped, const ColumnArray::Offsets & offsets, ColumnPtr & res_ptr)
{
using Result = ArrayAggregateResult<Element, aggregate_operation>;
using ResultType = ArrayAggregateResult<Element, aggregate_operation>;
using ColVecType = std::conditional_t<IsDecimalNumber<Element>, ColumnDecimal<Element>, ColumnVector<Element>>;
using ColVecResult = std::conditional_t<IsDecimalNumber<Result>, ColumnDecimal<Result>, ColumnVector<Result>>;
using ColVecResultType = std::conditional_t<IsDecimalNumber<ResultType>, ColumnDecimal<ResultType>, ColumnVector<ResultType>>;
/// For average and product of array we return Float64 as result, but we want to keep precision
/// so we convert to Float64 as last step, but intermediate sum is represented as result of sum operation
/// so we convert to Float64 as last step, but intermediate value is represented as result of sum operation
static constexpr bool is_average_or_product_operation = aggregate_operation == AggregateOperation::average ||
aggregate_operation == AggregateOperation::product;
using SummAggregationType = ArrayAggregateResult<Element, AggregateOperation::sum>;
using AggregationType = std::conditional_t<is_average_or_product_operation, SummAggregationType, Result>;
using AggregationType = std::conditional_t<is_average_or_product_operation, SummAggregationType, ResultType>;
const ColVecType * column = checkAndGetColumn<ColVecType>(&*mapped);
/// Constant case.
if (!column)
{
const ColumnConst * column_const = checkAndGetColumnConst<ColVecType>(&*mapped);
return false;
if (!column_const)
return false;
const auto & data = column->getData();
const AggregationType x = column_const->template getValue<Element>(); // NOLINT
const typename ColVecType::Container & data
= checkAndGetColumn<ColVecType>(&column_const->getDataColumn())->getData();
typename ColVecResult::MutablePtr res_column;
if constexpr (IsDecimalNumber<Element>)
{
res_column = ColVecResult::create(offsets.size(), data.getScale());
}
else
res_column = ColVecResult::create(offsets.size());
typename ColVecResult::Container & res = res_column->getData();
size_t pos = 0;
for (size_t i = 0; i < offsets.size(); ++i)
{
if constexpr (aggregate_operation == AggregateOperation::sum)
{
size_t array_size = offsets[i] - pos;
/// Just multiply the value by array size.
res[i] = x * array_size;
}
else if constexpr (aggregate_operation == AggregateOperation::min ||
aggregate_operation == AggregateOperation::max)
{
res[i] = x;
}
else if constexpr (aggregate_operation == AggregateOperation::average)
{
if constexpr (IsDecimalNumber<Element>)
{
res[i] = DecimalUtils::convertTo<Result>(x, data.getScale());
}
else
{
res[i] = x;
}
}
else if constexpr (aggregate_operation == AggregateOperation::product)
{
size_t array_size = offsets[i] - pos;
size_t num = array_size;
AggregationType product = x;
if constexpr (IsDecimalNumber<Element>)
{
using T = decltype(x.value);
T x_val = x.value;
for (i = 1; i < array_size; ++i)
{
T product_val = product.value;
while (common::mulOverflow(x_val, product_val, product.value))
{
x_val = x_val / DecimalUtils::scaleMultiplier<T>(data.getScale());
if (num == 1)
throw Exception("arrayProduct for decimal type overflow", ErrorCodes::DECIMAL_OVERFLOW);
--num;
}
}
res[i] = DecimalUtils::convertTo<Result>(product, data.getScale() * num);
}
else
{
for (i = 1; i < array_size; ++i)
product = product * x;
res[i] = product;
}
}
pos = offsets[i];
}
res_ptr = std::move(res_column);
return true;
}
const typename ColVecType::Container & data = column->getData();
typename ColVecResult::MutablePtr res_column;
typename ColVecResultType::MutablePtr res_column;
if constexpr (IsDecimalNumber<Element>)
res_column = ColVecResult::create(offsets.size(), data.getScale());
res_column = ColVecResultType::create(offsets.size(), data.getScale());
else
res_column = ColVecResult::create(offsets.size());
res_column = ColVecResultType::create(offsets.size());
typename ColVecResult::Container & res = res_column->getData();
typename ColVecResultType::Container & res = res_column->getData();
size_t pos = 0;
for (size_t i = 0; i < offsets.size(); ++i)
{
AggregationType s = 0;
AggregationType aggregate_value = 0;
/// Array is empty
if (offsets[i] == pos)
{
res[i] = s;
res[i] = aggregate_value;
continue;
}
size_t count = 1;
s = data[pos]; // NOLINT
aggregate_value = data[pos]; // NOLINT
++pos;
for (; pos < offsets[i]; ++pos)
@ -266,40 +182,36 @@ struct ArrayAggregateImpl
if constexpr (aggregate_operation == AggregateOperation::sum ||
aggregate_operation == AggregateOperation::average)
{
s += element;
aggregate_value += element;
}
else if constexpr (aggregate_operation == AggregateOperation::min)
{
if (element < s)
if (element < aggregate_value)
{
s = element;
aggregate_value = element;
}
}
else if constexpr (aggregate_operation == AggregateOperation::max)
{
if (element > s)
if (element > aggregate_value)
{
s = element;
aggregate_value = element;
}
}
else if constexpr (aggregate_operation == AggregateOperation::product)
{
if constexpr (IsDecimalNumber<Element>)
{
using T = decltype(s.value);
T s_val = s.value;
T element_val = element.value;
while (common::mulOverflow(s_val, element_val, s.value))
{
s_val = s_val / DecimalUtils::scaleMultiplier<T>(data.getScale());
if (count == 0)
throw Exception("arrayProduct for decimal type overflow", ErrorCodes::DECIMAL_OVERFLOW);
--count;
}
using AggregateValueDecimalUnderlyingValue = decltype(aggregate_value.value);
AggregateValueDecimalUnderlyingValue current_aggregate_value = aggregate_value.value;
AggregateValueDecimalUnderlyingValue element_value = static_cast<AggregateValueDecimalUnderlyingValue>(element.value);
if (common::mulOverflow(current_aggregate_value, element_value, aggregate_value.value))
throw Exception(ErrorCodes::DECIMAL_OVERFLOW, "Decimal math overflow");
}
else
{
s *= element;
aggregate_value *= element;
}
}
@ -310,21 +222,26 @@ struct ArrayAggregateImpl
{
if constexpr (IsDecimalNumber<Element>)
{
s = s / count;
res[i] = DecimalUtils::convertTo<Result>(s, data.getScale());
aggregate_value = aggregate_value / count;
res[i] = DecimalUtils::convertTo<ResultType>(aggregate_value, data.getScale());
}
else
{
res[i] = static_cast<Result>(s) / count;
res[i] = static_cast<ResultType>(aggregate_value) / count;
}
}
else if constexpr (aggregate_operation == AggregateOperation::product && IsDecimalNumber<Element>)
{
res[i] = DecimalUtils::convertTo<Result>(s, data.getScale() * count);
auto result_scale = data.getScale() * count;
if (unlikely(result_scale > DecimalUtils::max_precision<AggregationType>))
throw Exception(ErrorCodes::ARGUMENT_OUT_OF_BOUND, "Scale {} is out of bounds", result_scale);
res[i] = DecimalUtils::convertTo<ResultType>(aggregate_value, result_scale);
}
else
{
res[i] = s;
res[i] = aggregate_value;
}
}
@ -355,7 +272,7 @@ struct ArrayAggregateImpl
executeType<Decimal128>(mapped, offsets, res))
return res;
else
throw Exception("Unexpected column for arraySum: " + mapped->getName(), ErrorCodes::ILLEGAL_COLUMN);
throw Exception(ErrorCodes::ILLEGAL_COLUMN, "Unexpected column for arraySum: {}" + mapped->getName());
}
};

View File

@ -1,11 +1,13 @@
Array product 720
28.799999999999997 Float64
Array product with constant column
720 Float64
24 Float64
3.5 Float64
20 Float64
720
6 Float64
Array product with non constant column
24
0
6
720
24
0
6
Types of aggregation result array product

View File

@ -1,16 +1,20 @@
SELECT 'Array product ', (arrayProduct(array(1,2,3,4,5,6)));
select arrayProduct(array(1.0,2.0,3.0,4.8)) as k , toTypeName(k);
select arrayProduct(array(1,3.5)) as k , toTypeName(k);
SELECT arrayProduct([toDecimal32(2, 8), toDecimal32(10, 8)]) as a , toTypeName(a);
SELECT 'Array product with constant column';
SELECT arrayProduct([1,2,3,4,5,6]) as a, toTypeName(a);
SELECT arrayProduct(array(1.0,2.0,3.0,4.0)) as a, toTypeName(a);
SELECT arrayProduct(array(1,3.5)) as a, toTypeName(a);
SELECT arrayProduct([toDecimal64(1,8), toDecimal64(2,8), toDecimal64(3,8)]) as a, toTypeName(a);
SELECT 'Array product with non constant column';
DROP TABLE IF EXISTS test_aggregation;
CREATE TABLE test_aggregation (x Array(Int)) ENGINE=TinyLog;
INSERT INTO test_aggregation VALUES ([1,2,3,4,5,6]), ([]), ([1,2,3]);
INSERT INTO test_aggregation VALUES ([1,2,3,4]), ([]), ([1,2,3]);
SELECT arrayProduct(x) FROM test_aggregation;
DROP TABLE test_aggregation;
CREATE TABLE test_aggregation (x Array(Decimal64(8))) ENGINE=TinyLog;
INSERT INTO test_aggregation VALUES ([1,2,3,4,5,6]), ([]), ([1,2,3]);
INSERT INTO test_aggregation VALUES ([1,2,3,4]), ([]), ([1,2,3]);
SELECT arrayProduct(x) FROM test_aggregation;
DROP TABLE test_aggregation;