forked from NVIDIA/spark-rapids
-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Signed-off-by: sperlingxx <lovedreamf@gmail.com>
- Loading branch information
1 parent
be6abd3
commit 3b7b86e
Showing
5 changed files
with
205 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,34 @@ | ||
# Copyright (c) 2021, NVIDIA CORPORATION. | ||
# | ||
# Licensed under the Apache License, Version 2.0 (the "License"); | ||
# you may not use this file except in compliance with the License. | ||
# You may obtain a copy of the License at | ||
# | ||
# http://www.apache.org/licenses/LICENSE-2.0 | ||
# | ||
# Unless required by applicable law or agreed to in writing, software | ||
# distributed under the License is distributed on an "AS IS" BASIS, | ||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
# See the License for the specific language governing permissions and | ||
# limitations under the License. | ||
|
||
import pytest | ||
|
||
from asserts import assert_gpu_and_cpu_are_equal_collect | ||
from data_gen import * | ||
from pyspark.sql.types import * | ||
|
||
@pytest.mark.parametrize('data_gen', all_gen, ids=idfn) | ||
@pytest.mark.parametrize('size_of_null', ['true', 'false'], ids=idfn) | ||
def test_size_of_array(data_gen, size_of_null): | ||
gen = ArrayGen(data_gen) | ||
assert_gpu_and_cpu_are_equal_collect( | ||
lambda spark: unary_op_df(spark, gen).selectExpr('size(a)'), | ||
conf={'spark.sql.legacy.sizeOfNull': size_of_null}) | ||
|
||
@pytest.mark.parametrize('data_gen', map_gens_sample, ids=idfn) | ||
@pytest.mark.parametrize('size_of_null', ['true', 'false'], ids=idfn) | ||
def test_size_of_map(data_gen, size_of_null): | ||
assert_gpu_and_cpu_are_equal_collect( | ||
lambda spark: unary_op_df(spark, data_gen).selectExpr('size(a)'), | ||
conf={'spark.sql.legacy.sizeOfNull': size_of_null}) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
71 changes: 71 additions & 0 deletions
71
sql-plugin/src/main/scala/com/nvidia/spark/rapids/collectionOperations.scala
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,71 @@ | ||
/* | ||
* Copyright (c) 2021, NVIDIA CORPORATION. | ||
* | ||
* Licensed under the Apache License, Version 2.0 (the "License"); | ||
* you may not use this file except in compliance with the License. | ||
* You may obtain a copy of the License at | ||
* | ||
* http://www.apache.org/licenses/LICENSE-2.0 | ||
* | ||
* Unless required by applicable law or agreed to in writing, software | ||
* distributed under the License is distributed on an "AS IS" BASIS, | ||
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
* See the License for the specific language governing permissions and | ||
* limitations under the License. | ||
*/ | ||
|
||
package com.nvidia.spark.rapids | ||
|
||
import ai.rapids.cudf.ColumnVector | ||
|
||
import org.apache.spark.sql.catalyst.expressions.Expression | ||
import org.apache.spark.sql.types._ | ||
|
||
case class GpuSize(child: Expression, legacySizeOfNull: Boolean) | ||
extends GpuUnaryExpression { | ||
|
||
require(child.dataType.isInstanceOf[ArrayType] || child.dataType.isInstanceOf[MapType], | ||
s"The size function doesn't support the operand type ${child.dataType}") | ||
|
||
override def dataType: DataType = IntegerType | ||
override def nullable: Boolean = if (legacySizeOfNull) false else super.nullable | ||
|
||
override protected def doColumnar(input: GpuColumnVector): ColumnVector = { | ||
val inputBase = input.getBase | ||
if (inputBase.getRowCount == 0) { | ||
return GpuColumnVector.from(GpuScalar.from(0), 0, IntegerType).getBase | ||
} | ||
|
||
// Compute sizes of cuDF.ListType to get sizes of each ArrayData or MapData, considering | ||
// MapData is represented as List of Struct in terms of cuDF. | ||
// We compute list size via subtracting the offset of next element(row) to the current offset. | ||
val collectionSize = { | ||
// Here is a hack: using index -1 to fetch the offset column of list. | ||
// In terms of cuDF native, the offset is the first (index 0) child of list_column_view. | ||
// In JNI layer, we add 1 to the child index when fetching child column of ListType to keep | ||
// alignment. | ||
// So, in JVM layer, we have to use -1 as index to fetch the real first child of list_column. | ||
withResource(inputBase.getChildColumnView(-1)) { offset => | ||
withResource(offset.subVector(1)) { upBound => | ||
withResource(offset.subVector(0, offset.getRowCount.toInt - 1)) { lowBound => | ||
upBound.sub(lowBound) | ||
} | ||
} | ||
} | ||
} | ||
|
||
val nullScalar = if (legacySizeOfNull) { | ||
GpuScalar.from(-1) | ||
} else { | ||
GpuScalar.from(null, IntegerType) | ||
} | ||
|
||
withResource(collectionSize) { collectionSize => | ||
withResource(nullScalar) { nullScalar => | ||
withResource(inputBase.isNull) { inputIsNull => | ||
inputIsNull.ifElse(nullScalar, collectionSize) | ||
} | ||
} | ||
} | ||
} | ||
} |