|
| 1 | +// Licensed to the Apache Software Foundation (ASF) under one |
| 2 | +// or more contributor license agreements. See the NOTICE file |
| 3 | +// distributed with this work for additional information |
| 4 | +// regarding copyright ownership. The ASF licenses this file |
| 5 | +// to you under the Apache License, Version 2.0 (the |
| 6 | +// "License"); you may not use this file except in compliance |
| 7 | +// with the License. You may obtain a copy of the License at |
| 8 | +// |
| 9 | +// http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | +// |
| 11 | +// Unless required by applicable law or agreed to in writing, |
| 12 | +// software distributed under the License is distributed on an |
| 13 | +// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| 14 | +// KIND, either express or implied. See the License for the |
| 15 | +// specific language governing permissions and limitations |
| 16 | +// under the License. |
| 17 | + |
| 18 | +use arrow::array::ArrayRef; |
| 19 | +use arrow::datatypes::StringViewType; |
| 20 | +use arrow::util::bench_util::{ |
| 21 | + create_string_view_array_with_len, create_string_view_array_with_max_len, |
| 22 | +}; |
| 23 | +use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion}; |
| 24 | +use datafusion_physical_plan::aggregates::group_values::multi_group_by::bytes_view::ByteViewGroupValueBuilder; |
| 25 | +use datafusion_physical_plan::aggregates::group_values::multi_group_by::GroupColumn; |
| 26 | +use std::sync::Arc; |
| 27 | + |
| 28 | +const SIZES: [usize; 3] = [1_000, 10_000, 100_000]; |
| 29 | +const NULL_DENSITIES: [f32; 3] = [0.0, 0.1, 0.5]; |
| 30 | + |
| 31 | +fn bench_vectorized_append(c: &mut Criterion) { |
| 32 | + let mut group = c.benchmark_group("ByteViewGroupValueBuilder_vectorized_append"); |
| 33 | + |
| 34 | + for &size in &SIZES { |
| 35 | + let rows: Vec<usize> = (0..size).collect(); |
| 36 | + |
| 37 | + for &null_density in &NULL_DENSITIES { |
| 38 | + let input = create_string_view_array_with_len(size, null_density, 8, false); |
| 39 | + let input: ArrayRef = Arc::new(input); |
| 40 | + |
| 41 | + // vectorized_append |
| 42 | + let id = BenchmarkId::new( |
| 43 | + format!("inlined_null_{null_density:.1}_size_{size}"), |
| 44 | + "vectorized_append", |
| 45 | + ); |
| 46 | + group.bench_function(id, |b| { |
| 47 | + b.iter(|| { |
| 48 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 49 | + builder.vectorized_append(&input, &rows).unwrap(); |
| 50 | + }); |
| 51 | + }); |
| 52 | + |
| 53 | + // append_val |
| 54 | + let id = BenchmarkId::new( |
| 55 | + format!("inlined_null_{null_density:.1}_size_{size}"), |
| 56 | + "append_val", |
| 57 | + ); |
| 58 | + group.bench_function(id, |b| { |
| 59 | + b.iter(|| { |
| 60 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 61 | + for &i in &rows { |
| 62 | + builder.append_val(&input, i).unwrap(); |
| 63 | + } |
| 64 | + }); |
| 65 | + }); |
| 66 | + |
| 67 | + // vectorized_equal_to |
| 68 | + let id = BenchmarkId::new( |
| 69 | + format!("inlined_null_{null_density:.1}_size_{size}"), |
| 70 | + "vectorized_equal_to", |
| 71 | + ); |
| 72 | + group.bench_function(id, |b| { |
| 73 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 74 | + builder.vectorized_append(&input, &rows).unwrap(); |
| 75 | + let mut results = vec![true; size]; |
| 76 | + b.iter(|| { |
| 77 | + builder.vectorized_equal_to(&rows, &input, &rows, &mut results); |
| 78 | + }); |
| 79 | + }); |
| 80 | + } |
| 81 | + } |
| 82 | + |
| 83 | + for &size in &SIZES { |
| 84 | + let rows: Vec<usize> = (0..size).collect(); |
| 85 | + |
| 86 | + for &null_density in &NULL_DENSITIES { |
| 87 | + let scenario = "mixed"; |
| 88 | + let input = create_string_view_array_with_len(size, null_density, 64, true); |
| 89 | + let input: ArrayRef = Arc::new(input); |
| 90 | + |
| 91 | + // vectorized_append |
| 92 | + let id = BenchmarkId::new( |
| 93 | + format!("{scenario}_null_{null_density:.1}_size_{size}"), |
| 94 | + "vectorized_append", |
| 95 | + ); |
| 96 | + group.bench_function(id, |b| { |
| 97 | + b.iter(|| { |
| 98 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 99 | + builder.vectorized_append(&input, &rows).unwrap(); |
| 100 | + }); |
| 101 | + }); |
| 102 | + |
| 103 | + // append_val |
| 104 | + let id = BenchmarkId::new( |
| 105 | + format!("{scenario}_null_{null_density:.1}_size_{size}"), |
| 106 | + "append_val", |
| 107 | + ); |
| 108 | + group.bench_function(id, |b| { |
| 109 | + b.iter(|| { |
| 110 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 111 | + for &i in &rows { |
| 112 | + builder.append_val(&input, i).unwrap(); |
| 113 | + } |
| 114 | + }); |
| 115 | + }); |
| 116 | + |
| 117 | + // vectorized_equal_to |
| 118 | + let id = BenchmarkId::new( |
| 119 | + format!("{scenario}_null_{null_density:.1}_size_{size}"), |
| 120 | + "vectorized_equal_to", |
| 121 | + ); |
| 122 | + group.bench_function(id, |b| { |
| 123 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 124 | + builder.vectorized_append(&input, &rows).unwrap(); |
| 125 | + let mut results = vec![true; size]; |
| 126 | + b.iter(|| { |
| 127 | + builder.vectorized_equal_to(&rows, &input, &rows, &mut results); |
| 128 | + }); |
| 129 | + }); |
| 130 | + } |
| 131 | + } |
| 132 | + |
| 133 | + for &size in &SIZES { |
| 134 | + let rows: Vec<usize> = (0..size).collect(); |
| 135 | + |
| 136 | + for &null_density in &NULL_DENSITIES { |
| 137 | + let scenario = "random"; |
| 138 | + let input = create_string_view_array_with_max_len(size, null_density, 400); |
| 139 | + let input: ArrayRef = Arc::new(input); |
| 140 | + |
| 141 | + // vectorized_append |
| 142 | + let id = BenchmarkId::new( |
| 143 | + format!("{scenario}_null_{null_density:.1}_size_{size}"), |
| 144 | + "vectorized_append", |
| 145 | + ); |
| 146 | + group.bench_function(id, |b| { |
| 147 | + b.iter(|| { |
| 148 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 149 | + builder.vectorized_append(&input, &rows).unwrap(); |
| 150 | + }); |
| 151 | + }); |
| 152 | + |
| 153 | + // append_val |
| 154 | + let id = BenchmarkId::new( |
| 155 | + format!("{scenario}_null_{null_density:.1}_size_{size}"), |
| 156 | + "append_val", |
| 157 | + ); |
| 158 | + group.bench_function(id, |b| { |
| 159 | + b.iter(|| { |
| 160 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 161 | + for &i in &rows { |
| 162 | + builder.append_val(&input, i).unwrap(); |
| 163 | + } |
| 164 | + }); |
| 165 | + }); |
| 166 | + |
| 167 | + // vectorized_equal_to |
| 168 | + let id = BenchmarkId::new( |
| 169 | + format!("{scenario}_null_{null_density:.1}_size_{size}"), |
| 170 | + "vectorized_equal_to", |
| 171 | + ); |
| 172 | + group.bench_function(id, |b| { |
| 173 | + let mut builder = ByteViewGroupValueBuilder::<StringViewType>::new(); |
| 174 | + builder.vectorized_append(&input, &rows).unwrap(); |
| 175 | + let mut results = vec![true; size]; |
| 176 | + b.iter(|| { |
| 177 | + builder.vectorized_equal_to(&rows, &input, &rows, &mut results); |
| 178 | + }); |
| 179 | + }); |
| 180 | + } |
| 181 | + } |
| 182 | + |
| 183 | + group.finish(); |
| 184 | +} |
| 185 | + |
| 186 | +criterion_group!(benches, bench_vectorized_append); |
| 187 | +criterion_main!(benches); |
0 commit comments