This is an automated email from the ASF dual-hosted git repository.
alamb pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/datafusion.git
The following commit(s) were added to refs/heads/main by this push:
new 55a1459b12 benches: add lower benches (#12152)
55a1459b12 is described below
commit 55a1459b123c0bb16e2cbe41cc59cf3582115932
Author: Trent Hauck <[email protected]>
AuthorDate: Mon Aug 26 12:19:27 2024 -0700
benches: add lower benches (#12152)
---
datafusion/functions/benches/lower.rs | 92 ++++++++++++++++++++++++++++++++++-
1 file changed, 90 insertions(+), 2 deletions(-)
diff --git a/datafusion/functions/benches/lower.rs
b/datafusion/functions/benches/lower.rs
index fa963f174e..934c1c6bd1 100644
--- a/datafusion/functions/benches/lower.rs
+++ b/datafusion/functions/benches/lower.rs
@@ -17,8 +17,10 @@
extern crate criterion;
-use arrow::array::{ArrayRef, StringArray};
-use arrow::util::bench_util::create_string_array_with_len;
+use arrow::array::{ArrayRef, StringArray, StringViewBuilder};
+use arrow::util::bench_util::{
+ create_string_array_with_len, create_string_view_array_with_len,
+};
use criterion::{black_box, criterion_group, criterion_main, Criterion};
use datafusion_expr::ColumnarValue;
use datafusion_functions::string;
@@ -65,6 +67,58 @@ fn create_args3(size: usize) -> Vec<ColumnarValue> {
vec![ColumnarValue::Array(array)]
}
+/// Create an array of args containing StringViews, where all the values in the
+/// StringViews are ASCII.
+/// * `size` - the length of the StringViews, and
+/// * `str_len` - the length of the strings within the array.
+/// * `null_density` - the density of null values in the array.
+/// * `mixed` - whether the array is mixed between inlined and referenced
strings.
+fn create_args4(
+ size: usize,
+ str_len: usize,
+ null_density: f32,
+ mixed: bool,
+) -> Vec<ColumnarValue> {
+ let array = Arc::new(create_string_view_array_with_len(
+ size,
+ null_density,
+ str_len,
+ mixed,
+ ));
+
+ vec![ColumnarValue::Array(array)]
+}
+
+/// Create an array of args containing a StringViewArray, where some of the
values in the
+/// array are non-ASCII.
+/// * `size` - the length of the StringArray, and
+/// * `non_ascii_density` - the density of non-ASCII values in the array.
+/// * `null_density` - the density of null values in the array.
+fn create_args5(
+ size: usize,
+ non_ascii_density: f32,
+ null_density: f32,
+) -> Vec<ColumnarValue> {
+ let mut string_view_builder = StringViewBuilder::with_capacity(size);
+ for _ in 0..size {
+ // sample null_density to determine if the value should be null
+ if rand::random::<f32>() < null_density {
+ string_view_builder.append_null();
+ continue;
+ }
+
+ // sample non_ascii_density to determine if the value should be
non-ASCII
+ if rand::random::<f32>() < non_ascii_density {
+ string_view_builder.append_value("农历新年农历新年农历新年农历新年农历新年");
+ } else {
+ string_view_builder.append_value("DATAFUSIONDATAFUSIONDATAFUSION");
+ }
+ }
+
+ let array = Arc::new(string_view_builder.finish()) as ArrayRef;
+ vec![ColumnarValue::Array(array)]
+}
+
fn criterion_benchmark(c: &mut Criterion) {
let lower = string::lower();
for size in [1024, 4096, 8192] {
@@ -85,6 +139,40 @@ fn criterion_benchmark(c: &mut Criterion) {
|b| b.iter(|| black_box(lower.invoke(&args))),
);
}
+
+ let sizes = [4096, 8192];
+ let str_lens = [10, 64, 128];
+ let mixes = [true, false];
+ let null_densities = [0.0f32, 0.1f32];
+
+ for null_density in &null_densities {
+ for &mixed in &mixes {
+ for &str_len in &str_lens {
+ for &size in &sizes {
+ let args = create_args4(size, str_len, *null_density,
mixed);
+ c.bench_function(
+ &format!("lower_all_values_are_ascii_string_views: size: {},
str_len: {}, null_density: {}, mixed: {}",
+ size, str_len, null_density, mixed),
+ |b| b.iter(|| black_box(lower.invoke(&args))),
+ );
+
+ let args = create_args4(size, str_len, *null_density,
mixed);
+ c.bench_function(
+ &format!("lower_all_values_are_ascii_string_views: size: {},
str_len: {}, null_density: {}, mixed: {}",
+ size, str_len, null_density, mixed),
+ |b| b.iter(|| black_box(lower.invoke(&args))),
+ );
+
+ let args = create_args5(size, 0.1, *null_density);
+ c.bench_function(
+ &format!("lower_some_values_are_nonascii_string_views: size: {},
str_len: {}, non_ascii_density: {}, null_density: {}, mixed: {}",
+ size, str_len, 0.1, null_density, mixed),
+ |b| b.iter(|| black_box(lower.invoke(&args))),
+ );
+ }
+ }
+ }
+ }
}
criterion_group!(benches, criterion_benchmark);
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
