-
Notifications
You must be signed in to change notification settings - Fork 912
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge branch 'branch-24.12' into fea-strings-find-re
- Loading branch information
Showing
36 changed files
with
1,209 additions
and
181 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
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -50,6 +50,7 @@ jobs: | |
| test_java: ${{ steps.changed-files.outputs.java_any_changed == 'true' }} | ||
| test_notebooks: ${{ steps.changed-files.outputs.notebooks_any_changed == 'true' }} | ||
| test_python: ${{ steps.changed-files.outputs.python_any_changed == 'true' }} | ||
| test_cudf_pandas: ${{ steps.changed-files.outputs.cudf_pandas_any_changed == 'true' }} | ||
| steps: | ||
| - name: Get PR info | ||
| id: get-pr-info | ||
|
|
@@ -82,6 +83,7 @@ jobs: | |
| - '!java/**' | ||
| - '!notebooks/**' | ||
| - '!python/**' | ||
| - '!ci/cudf_pandas_scripts/**' | ||
| java: | ||
| - '**' | ||
| - '!CONTRIBUTING.md' | ||
|
|
@@ -90,11 +92,13 @@ jobs: | |
| - '!img/**' | ||
| - '!notebooks/**' | ||
| - '!python/**' | ||
| - '!ci/cudf_pandas_scripts/**' | ||
| notebooks: | ||
| - '**' | ||
| - '!CONTRIBUTING.md' | ||
| - '!README.md' | ||
| - '!java/**' | ||
| - '!ci/cudf_pandas_scripts/**' | ||
| python: | ||
| - '**' | ||
| - '!CONTRIBUTING.md' | ||
|
|
@@ -103,6 +107,16 @@ jobs: | |
| - '!img/**' | ||
| - '!java/**' | ||
| - '!notebooks/**' | ||
| - '!ci/cudf_pandas_scripts/**' | ||
| cudf_pandas: | ||
| - '**' | ||
| - 'ci/cudf_pandas_scripts/**' | ||
| - '!CONTRIBUTING.md' | ||
| - '!README.md' | ||
| - '!docs/**' | ||
| - '!img/**' | ||
| - '!java/**' | ||
| - '!notebooks/**' | ||
| checks: | ||
| secrets: inherit | ||
| uses: rapidsai/shared-workflows/.github/workflows/[email protected] | ||
|
|
@@ -248,7 +262,7 @@ jobs: | |
| cudf-polars-polars-tests: | ||
| needs: wheel-build-cudf-polars | ||
| secrets: inherit | ||
| uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@branch-24.10 | ||
| uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@branch-24.12 | ||
| with: | ||
| # This selects "ARCH=amd64 + the latest supported Python + CUDA". | ||
| matrix_filter: map(select(.ARCH == "amd64")) | group_by(.CUDA_VER|split(".")|map(tonumber)|.[0]) | map(max_by([(.PY_VER|split(".")|map(tonumber)), (.CUDA_VER|split(".")|map(tonumber))])) | ||
|
|
@@ -289,7 +303,7 @@ jobs: | |
| needs: [wheel-build-cudf, changed-files] | ||
| secrets: inherit | ||
| uses: rapidsai/shared-workflows/.github/workflows/[email protected] | ||
| if: needs.changed-files.outputs.test_python == 'true' | ||
| if: needs.changed-files.outputs.test_python == 'true' || needs.changed-files.outputs.test_cudf_pandas == 'true' | ||
| with: | ||
| # This selects "ARCH=amd64 + the latest supported Python + CUDA". | ||
| matrix_filter: map(select(.ARCH == "amd64")) | group_by(.CUDA_VER|split(".")|map(tonumber)|.[0]) | map(max_by([(.PY_VER|split(".")|map(tonumber)), (.CUDA_VER|split(".")|map(tonumber))])) | ||
|
|
@@ -300,7 +314,7 @@ jobs: | |
| needs: [wheel-build-cudf, changed-files] | ||
| secrets: inherit | ||
| uses: rapidsai/shared-workflows/.github/workflows/[email protected] | ||
| if: needs.changed-files.outputs.test_python == 'true' | ||
| if: needs.changed-files.outputs.test_python == 'true' || needs.changed-files.outputs.test_cudf_pandas == 'true' | ||
| with: | ||
| # This selects "ARCH=amd64 + the latest supported Python + CUDA". | ||
| matrix_filter: map(select(.ARCH == "amd64")) | group_by(.CUDA_VER|split(".")|map(tonumber)|.[0]) | map(max_by([(.PY_VER|split(".")|map(tonumber)), (.CUDA_VER|split(".")|map(tonumber))])) | ||
|
|
||
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
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 +1,123 @@ | ||
| /* | ||
| * Copyright (c) 2024, 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. | ||
| */ | ||
|
|
||
| #include <cudf_test/column_wrapper.hpp> | ||
|
|
||
| #include <cudf/detail/tdigest/tdigest.hpp> | ||
| #include <cudf/utilities/default_stream.hpp> | ||
|
|
||
| #include <rmm/exec_policy.hpp> | ||
|
|
||
| #include <cuda/functional> | ||
| #include <thrust/copy.h> | ||
| #include <thrust/execution_policy.h> | ||
|
|
||
| #include <nvbench/nvbench.cuh> | ||
|
|
||
| void bm_tdigest_merge(nvbench::state& state) | ||
| { | ||
| auto const num_tdigests = static_cast<cudf::size_type>(state.get_int64("num_tdigests")); | ||
| auto const tdigest_size = static_cast<cudf::size_type>(state.get_int64("tdigest_size")); | ||
| auto const tdigests_per_group = | ||
| static_cast<cudf::size_type>(state.get_int64("tdigests_per_group")); | ||
| auto const max_centroids = static_cast<cudf::size_type>(state.get_int64("max_centroids")); | ||
| auto const num_groups = num_tdigests / tdigests_per_group; | ||
| auto const total_centroids = num_tdigests * tdigest_size; | ||
|
|
||
| auto stream = cudf::get_default_stream(); | ||
| auto mr = rmm::mr::get_current_device_resource(); | ||
|
|
||
| constexpr int base_value = 5; | ||
|
|
||
| // construct inner means/weights | ||
| auto val_iter = cudf::detail::make_counting_transform_iterator( | ||
| 0, cuda::proclaim_return_type<double>([tdigest_size](cudf::size_type i) { | ||
| return static_cast<double>(base_value + (i % tdigest_size)); | ||
| })); | ||
| auto one_iter = thrust::make_constant_iterator(1); | ||
| cudf::test::fixed_width_column_wrapper<double> means(val_iter, val_iter + total_centroids); | ||
| cudf::test::fixed_width_column_wrapper<double> weights(one_iter, one_iter + total_centroids); | ||
| std::vector<std::unique_ptr<cudf::column>> inner_struct_children; | ||
| inner_struct_children.push_back(means.release()); | ||
| inner_struct_children.push_back(weights.release()); | ||
| cudf::test::structs_column_wrapper inner_struct(std::move(inner_struct_children)); | ||
|
|
||
| // construct the tdigest lists themselves | ||
| auto offset_iter = cudf::detail::make_counting_transform_iterator( | ||
| 0, cuda::proclaim_return_type<cudf::size_type>([tdigest_size](cudf::size_type i) { | ||
| return i * tdigest_size; | ||
| })); | ||
| cudf::test::fixed_width_column_wrapper<int> offsets(offset_iter, offset_iter + num_tdigests + 1); | ||
| auto list_col = cudf::make_lists_column( | ||
| num_tdigests, offsets.release(), inner_struct.release(), 0, {}, stream, mr); | ||
|
|
||
| // min and max columns | ||
| auto min_iter = thrust::make_constant_iterator(base_value); | ||
| auto max_iter = thrust::make_constant_iterator(base_value + (tdigest_size - 1)); | ||
| cudf::test::fixed_width_column_wrapper<double> mins(min_iter, min_iter + num_tdigests); | ||
| cudf::test::fixed_width_column_wrapper<double> maxes(max_iter, max_iter + num_tdigests); | ||
|
|
||
| // assemble the whole thing | ||
| std::vector<std::unique_ptr<cudf::column>> tdigest_children; | ||
| tdigest_children.push_back(std::move(list_col)); | ||
| tdigest_children.push_back(mins.release()); | ||
| tdigest_children.push_back(maxes.release()); | ||
| cudf::test::structs_column_wrapper tdigest(std::move(tdigest_children)); | ||
|
|
||
| rmm::device_uvector<cudf::size_type> group_offsets(num_groups + 1, stream, mr); | ||
| rmm::device_uvector<cudf::size_type> group_labels(num_tdigests, stream, mr); | ||
| auto group_offset_iter = cudf::detail::make_counting_transform_iterator( | ||
| 0, | ||
| cuda::proclaim_return_type<cudf::size_type>( | ||
| [tdigests_per_group] __device__(cudf::size_type i) { return i * tdigests_per_group; })); | ||
| thrust::copy(rmm::exec_policy_nosync(stream, mr), | ||
| group_offset_iter, | ||
| group_offset_iter + num_groups + 1, | ||
| group_offsets.begin()); | ||
| auto group_label_iter = cudf::detail::make_counting_transform_iterator( | ||
| 0, | ||
| cuda::proclaim_return_type<cudf::size_type>( | ||
| [tdigests_per_group] __device__(cudf::size_type i) { return i / tdigests_per_group; })); | ||
| thrust::copy(rmm::exec_policy_nosync(stream, mr), | ||
| group_label_iter, | ||
| group_label_iter + num_tdigests, | ||
| group_labels.begin()); | ||
|
|
||
| state.add_element_count(total_centroids); | ||
|
|
||
| state.set_cuda_stream(nvbench::make_cuda_stream_view(stream.value())); | ||
| state.exec(nvbench::exec_tag::timer | nvbench::exec_tag::sync, | ||
| [&](nvbench::launch& launch, auto& timer) { | ||
| timer.start(); | ||
| auto result = cudf::tdigest::detail::group_merge_tdigest( | ||
| tdigest, group_offsets, group_labels, num_groups, max_centroids, stream, mr); | ||
| timer.stop(); | ||
| }); | ||
| } | ||
|
|
||
| NVBENCH_BENCH(bm_tdigest_merge) | ||
| .set_name("TDigest many tiny groups") | ||
| .add_int64_axis("num_tdigests", {500'000}) | ||
| .add_int64_axis("tdigest_size", {1, 1000}) | ||
| .add_int64_axis("tdigests_per_group", {1}) | ||
| .add_int64_axis("max_centroids", {10000, 1000}); | ||
|
|
||
| NVBENCH_BENCH(bm_tdigest_merge) | ||
| .set_name("TDigest many small groups") | ||
| .add_int64_axis("num_tdigests", {500'000}) | ||
| .add_int64_axis("tdigest_size", {1, 1000}) | ||
| .add_int64_axis("tdigests_per_group", {3}) | ||
| .add_int64_axis("max_centroids", {10000, 1000}); |
Oops, something went wrong.