BenMagyar commented on code in PR #50534:
URL: https://github.com/apache/arrow/pull/50534#discussion_r3678096773
##########
cpp/src/arrow/compute/kernels/vector_run_end_encode.cc:
##########
@@ -470,6 +536,56 @@ struct RunEndDecodeExec {
return Status::Invalid("Invalid run end type: ",
*ree_type->run_end_type());
}
+ template <typename RunEndType>
+ static Status DoExecNested(KernelContext* ctx, const ArraySpan& input_array,
+ ExecResult* result) {
+ using RunEndCType = typename RunEndType::c_type;
+ const auto& input_values = arrow::ree_util::ValuesArray(input_array);
+
+ ARROW_ASSIGN_OR_RAISE(
+ auto output_builder,
+ MakeBuilderExactIndex(input_values.type->GetSharedPtr(),
ctx->memory_pool()));
+ RETURN_NOT_OK(output_builder->Reserve(input_array.length));
+
+ const arrow::ree_util::RunEndEncodedArraySpan<RunEndCType> ree_array_span(
+ input_array);
+ if (input_array.length > 0) {
+ for (auto it = ree_array_span.begin(); !it.is_end(ree_array_span); ++it)
{
+ const int64_t physical_index = it.index_into_array();
+ const int64_t run_length = it.run_length();
+ if (input_values.IsNull(physical_index)) {
+ RETURN_NOT_OK(output_builder->AppendNulls(run_length));
+ continue;
+ }
+ for (int64_t i = 0; i < run_length; ++i) {
+ RETURN_NOT_OK(
+ output_builder->AppendArraySlice(input_values, physical_index,
1));
+ }
Review Comment:
I ran some some mini benchmarks to test this:
| Value type | Run length 1 | Run length 64 | Run length 4096 | Run length
16384 |
|---|---:|---:|---:|---:|
| `List<Int32>[4]` | 32.5x slower | 12.1x slower | 11.5x slower | 11.6x
slower |
| `List<Int32>[64]` | 126.3x slower | 116.3x slower | 115.9x slower | 115.2x
slower |
| `LargeList<Int32>[8]` | 41.9x slower | 22.3x slower | 21.8x slower | 22.1x
slower |
| `ListView<Int32>[8]` | 41.7x slower | 22.1x slower | 21.7x slower | 21.7x
slower |
| `LargeListView<Int32>[8]` | 42.3x slower | 22.2x slower | 21.9x slower |
21.6x slower |
| `FixedSizeList<Int32>[8]` | 53.4x slower | 26.2x slower | 25.7x slower |
25.6x slower |
| `Map<Utf8, Int32>[4]` | 130.5x slower | 107.9x slower | 109.6x slower |
107.0x slower |
| `Struct mixed` | 58.9x slower | 16.5x slower | 15.6x slower | 15.7x slower
|
| Empty `List` | 24.2x slower | 0.75x | 0.35x | 0.35x |
| `List<Int32>[1]` | 23.7x slower | 3.5x slower | 3.1x slower | 3.1x slower |
| Empty `Map` | 28.0x slower | 1.0x | 0.52x | 0.51x |
| `Struct<Int32>` | 34.9x slower | 1.4x slower | 0.84x | 0.77x |
| Empty `Struct` | 30.6x slower | 1.1x slower | 0.56x | 0.56x |
|
##########
cpp/src/arrow/compute/kernels/vector_run_end_encode.cc:
##########
@@ -470,6 +536,56 @@ struct RunEndDecodeExec {
return Status::Invalid("Invalid run end type: ",
*ree_type->run_end_type());
}
+ template <typename RunEndType>
+ static Status DoExecNested(KernelContext* ctx, const ArraySpan& input_array,
+ ExecResult* result) {
+ using RunEndCType = typename RunEndType::c_type;
+ const auto& input_values = arrow::ree_util::ValuesArray(input_array);
+
+ ARROW_ASSIGN_OR_RAISE(
+ auto output_builder,
+ MakeBuilderExactIndex(input_values.type->GetSharedPtr(),
ctx->memory_pool()));
+ RETURN_NOT_OK(output_builder->Reserve(input_array.length));
+
+ const arrow::ree_util::RunEndEncodedArraySpan<RunEndCType> ree_array_span(
+ input_array);
+ if (input_array.length > 0) {
+ for (auto it = ree_array_span.begin(); !it.is_end(ree_array_span); ++it)
{
+ const int64_t physical_index = it.index_into_array();
+ const int64_t run_length = it.run_length();
+ if (input_values.IsNull(physical_index)) {
+ RETURN_NOT_OK(output_builder->AppendNulls(run_length));
+ continue;
+ }
+ for (int64_t i = 0; i < run_length; ++i) {
+ RETURN_NOT_OK(
+ output_builder->AppendArraySlice(input_values, physical_index,
1));
+ }
Review Comment:
I ran some some mini benchmarks to test this:
| Value type | Run length 1 | Run length 64 | Run length 4096 | Run length
16384 |
|---|---:|---:|---:|---:|
| `List<Int32>[4]` | 32.5x slower | 12.1x slower | 11.5x slower | 11.6x
slower |
| `List<Int32>[64]` | 126.3x slower | 116.3x slower | 115.9x slower | 115.2x
slower |
| `LargeList<Int32>[8]` | 41.9x slower | 22.3x slower | 21.8x slower | 22.1x
slower |
| `ListView<Int32>[8]` | 41.7x slower | 22.1x slower | 21.7x slower | 21.7x
slower |
| `LargeListView<Int32>[8]` | 42.3x slower | 22.2x slower | 21.9x slower |
21.6x slower |
| `FixedSizeList<Int32>[8]` | 53.4x slower | 26.2x slower | 25.7x slower |
25.6x slower |
| `Map<Utf8, Int32>[4]` | 130.5x slower | 107.9x slower | 109.6x slower |
107.0x slower |
| `Struct mixed` | 58.9x slower | 16.5x slower | 15.6x slower | 15.7x slower
|
| Empty `List` | 24.2x slower | 0.75x | 0.35x | 0.35x |
| `List<Int32>[1]` | 23.7x slower | 3.5x slower | 3.1x slower | 3.1x slower |
| Empty `Map` | 28.0x slower | 1.0x | 0.52x | 0.51x |
| `Struct<Int32>` | 34.9x slower | 1.4x slower | 0.84x | 0.77x |
| Empty `Struct` | 30.6x slower | 1.1x slower | 0.56x | 0.56x |
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
