Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add cardinality control for groupby benchs with flat types #15134

Merged
merged 8 commits into from
Mar 8, 2024
Merged

Add cardinality control for groupby benchs with flat types #15134

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
29d3e23
Add groupby_max_cardinality bench
PointKernel Feb 24, 2024
45f8073
Minor cleanups
PointKernel Feb 24, 2024
5893f22
Merge remote-tracking branch 'upstream/branch-24.04' into groupby-ben…
PointKernel Feb 29, 2024
f6a13f4
Merge remote-tracking branch 'upstream/branch-24.04' into groupby-ben…
PointKernel Mar 4, 2024
77e53cb
Add cardinality axis to the normal groupby max bench
PointKernel Mar 4, 2024
30f53e8
Add cardinality control for group_rank and group_nunique
PointKernel Mar 4, 2024
da8477a
Merge remote-tracking branch 'upstream/branch-24.04' into groupby-ben…
PointKernel Mar 4, 2024
99138c7
Merge branch 'branch-24.04' into groupby-bench-cardinality
PointKernel Mar 4, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
53 changes: 42 additions & 11 deletions cpp/benchmarks/groupby/group_max.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,25 +22,30 @@
#include <nvbench/nvbench.cuh>

template <typename Type>
void bench_groupby_max(nvbench::state& state, nvbench::type_list<Type>)
void groupby_max_helper(nvbench::state& state,
cudf::size_type num_rows,
cudf::size_type cardinality,
double null_probability)
{
auto const size = static_cast<cudf::size_type>(state.get_int64("num_rows"));

auto const keys = [&] {
data_profile const profile = data_profile_builder().cardinality(0).no_validity().distribution(
cudf::type_to_id<int32_t>(), distribution_id::UNIFORM, 0, 100);
return create_random_column(cudf::type_to_id<int32_t>(), row_count{size}, profile);
data_profile const profile =
data_profile_builder()
.cardinality(cardinality)
.no_validity()
.distribution(cudf::type_to_id<int32_t>(), distribution_id::UNIFORM, 0, num_rows);
return create_random_column(cudf::type_to_id<int32_t>(), row_count{num_rows}, profile);
}();

auto const vals = [&] {
auto builder = data_profile_builder().cardinality(0).distribution(
cudf::type_to_id<Type>(), distribution_id::UNIFORM, 0, 1000);
if (const auto null_freq = state.get_float64("null_probability"); null_freq > 0) {
builder.null_probability(null_freq);
cudf::type_to_id<Type>(), distribution_id::UNIFORM, 0, num_rows);
if (null_probability > 0) {
builder.null_probability(null_probability);
} else {
builder.no_validity();
}
return create_random_column(cudf::type_to_id<Type>(), row_count{size}, data_profile{builder});
return create_random_column(
cudf::type_to_id<Type>(), row_count{num_rows}, data_profile{builder});
}();

auto keys_view = keys->view();
Expand All @@ -55,13 +60,39 @@ void bench_groupby_max(nvbench::state& state, nvbench::type_list<Type>)
state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
state.exec(nvbench::exec_tag::sync,
[&](nvbench::launch& launch) { auto const result = gb_obj.aggregate(requests); });

auto const elapsed_time = state.get_summary("nv/cold/time/gpu/mean").get_float64("value");
state.add_element_count(static_cast<double>(num_rows) / elapsed_time / 1'000'000., "Mrows/s");
state.add_buffer_size(
mem_stats_logger.peak_memory_usage(), "peak_memory_usage", "peak_memory_usage");
}

template <typename Type>
void bench_groupby_max(nvbench::state& state, nvbench::type_list<Type>)
{
auto const cardinality = static_cast<cudf::size_type>(state.get_int64("cardinality"));
auto const num_rows = static_cast<cudf::size_type>(state.get_int64("num_rows"));
auto const null_probability = state.get_float64("null_probability");

groupby_max_helper<Type>(state, num_rows, cardinality, null_probability);
}

template <typename Type>
void bench_groupby_max_cardinality(nvbench::state& state, nvbench::type_list<Type>)
{
auto constexpr num_rows = 20'000'000;
auto constexpr null_probability = 0.;
auto const cardinality = static_cast<cudf::size_type>(state.get_int64("cardinality"));

groupby_max_helper<Type>(state, num_rows, cardinality, null_probability);
}

NVBENCH_BENCH_TYPES(bench_groupby_max,
NVBENCH_TYPE_AXES(nvbench::type_list<int32_t, int64_t, float, double>))
.set_name("groupby_max")
.add_int64_axis("cardinality", {0})
.add_int64_power_of_two_axis("num_rows", {12, 18, 24})
.add_float64_axis("null_probability", {0, 0.1, 0.9});

NVBENCH_BENCH_TYPES(bench_groupby_max_cardinality, NVBENCH_TYPE_AXES(nvbench::type_list<int32_t>))
.set_name("groupby_max_cardinality")
.add_int64_axis("cardinality", {10, 20, 50, 100, 1'000, 10'000, 100'000, 1'000'000, 10'000'000});
19 changes: 13 additions & 6 deletions cpp/benchmarks/groupby/group_nunique.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* Copyright (c) 2022-2023, NVIDIA CORPORATION.
* Copyright (c) 2022-2024, NVIDIA CORPORATION.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
Expand Down Expand Up @@ -39,17 +39,23 @@ auto make_aggregation_request_vector(cudf::column_view const& values, Args&&...
template <typename Type>
void bench_groupby_nunique(nvbench::state& state, nvbench::type_list<Type>)
{
auto const size = static_cast<cudf::size_type>(state.get_int64("num_rows"));
auto const size = static_cast<cudf::size_type>(state.get_int64("num_rows"));
auto const cardinality = static_cast<cudf::size_type>(state.get_int64("cardinality"));

auto const keys = [&] {
data_profile profile = data_profile_builder().cardinality(0).no_validity().distribution(
cudf::type_to_id<int32_t>(), distribution_id::UNIFORM, 0, 100);
data_profile profile =
data_profile_builder()
.cardinality(cardinality)
.no_validity()
.distribution(cudf::type_to_id<int32_t>(), distribution_id::UNIFORM, 0, size);
Comment on lines +46 to +50
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code to create keys and values here seem to be very similar (the same) as for creating those of groupby max. Can we further extract them into a common function? Like create_keys_values()?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The duplicate part is just data profile construction. Though we repeat those 5 lines of code each time, it's just one API call. It's not worth creating a helper function wrapping one single call IMO.

return create_random_column(cudf::type_to_id<int32_t>(), row_count{size}, profile);
}();

auto const vals = [&] {
data_profile profile = data_profile_builder().cardinality(0).distribution(
cudf::type_to_id<Type>(), distribution_id::UNIFORM, 0, 1000);
data_profile profile =
data_profile_builder()
.cardinality(cardinality)
.distribution(cudf::type_to_id<Type>(), distribution_id::UNIFORM, 0, size);
if (const auto null_freq = state.get_float64("null_probability"); null_freq > 0) {
profile.set_null_probability(null_freq);
} else {
Expand All @@ -71,4 +77,5 @@ void bench_groupby_nunique(nvbench::state& state, nvbench::type_list<Type>)
NVBENCH_BENCH_TYPES(bench_groupby_nunique, NVBENCH_TYPE_AXES(nvbench::type_list<int32_t, int64_t>))
.set_name("groupby_nunique")
.add_int64_power_of_two_axis("num_rows", {12, 16, 20, 24})
.add_int64_axis("cardinality", {0})
.add_float64_axis("null_probability", {0, 0.5});
12 changes: 7 additions & 5 deletions cpp/benchmarks/groupby/group_rank.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* Copyright (c) 2022-2023, NVIDIA CORPORATION.
* Copyright (c) 2022-2024, NVIDIA CORPORATION.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
Expand Down Expand Up @@ -31,10 +31,12 @@ static void nvbench_groupby_rank(nvbench::state& state,

bool const is_sorted = state.get_int64("is_sorted");
cudf::size_type const column_size = state.get_int64("data_size");
constexpr int num_groups = 100;
auto const cardinality = static_cast<cudf::size_type>(state.get_int64("cardinality"));

data_profile const profile = data_profile_builder().cardinality(0).no_validity().distribution(
dtype, distribution_id::UNIFORM, 0, num_groups);
data_profile const profile = data_profile_builder()
.cardinality(cardinality)
.no_validity()
.distribution(dtype, distribution_id::UNIFORM, 0, column_size);

auto source_table = create_random_table({dtype, dtype}, row_count{column_size}, profile);

Expand Down Expand Up @@ -100,5 +102,5 @@ NVBENCH_BENCH_TYPES(nvbench_groupby_rank, NVBENCH_TYPE_AXES(methods))
10000000, // 10M
100000000, // 100M
})

.add_int64_axis("cardinality", {0})
.add_int64_axis("is_sorted", {0, 1});
Loading