Add string_dictionary benches for row format (apache#2677)

tustvold · Oct 3, 2022 · 8046e07 · 8046e07
1 parent da8f742
commit 8046e07
Showing 1 changed file with 66 additions and 1 deletion.
diff --git a/arrow/benches/row_format.rs b/arrow/benches/row_format.rs
@@ -22,9 +22,10 @@ extern crate core;
 use arrow::array::ArrayRef;
 use arrow::datatypes::{DataType, Int64Type, UInt64Type};
 use arrow::row::{RowConverter, SortField};
-use arrow::util::bench_util::{create_primitive_array, create_string_array_with_len};
+use arrow::util::bench_util::{create_primitive_array, create_string_array_with_len, create_string_dict_array};
 use criterion::{black_box, Criterion};
 use std::sync::Arc;
+use arrow_array::types::Int32Type;
 
 fn row_bench(c: &mut Criterion) {
     let cols = vec![Arc::new(create_primitive_array::<UInt64Type>(4096, 0.)) as ArrayRef];
@@ -85,6 +86,46 @@ fn row_bench(c: &mut Criterion) {
         });
     });
 
+    let cols =
+        vec![Arc::new(create_string_dict_array::<Int32Type>(4096, 0., 100)) as ArrayRef];
+
+    c.bench_function("row_batch 4096 string_dictionary(10, 0)", |b| {
+        b.iter(|| {
+            let mut converter = RowConverter::new(vec![SortField::new(DataType::Utf8)]);
+            black_box(converter.convert_columns(&cols))
+        });
+    });
+
+    let cols =
+        vec![Arc::new(create_string_dict_array::<Int32Type>(4096, 0., 100)) as ArrayRef];
+
+    c.bench_function("row_batch 4096 string_dictionary(30, 0)", |b| {
+        b.iter(|| {
+            let mut converter = RowConverter::new(vec![SortField::new(DataType::Utf8)]);
+            black_box(converter.convert_columns(&cols))
+        });
+    });
+
+    let cols =
+        vec![Arc::new(create_string_dict_array::<Int32Type>(4096, 0., 100)) as ArrayRef];
+
+    c.bench_function("row_batch 4096 string_dictionary(100, 0)", |b| {
+        b.iter(|| {
+            let mut converter = RowConverter::new(vec![SortField::new(DataType::Utf8)]);
+            black_box(converter.convert_columns(&cols))
+        });
+    });
+
+    let cols =
+        vec![Arc::new(create_string_dict_array::<Int32Type>(4096, 0.5, 100)) as ArrayRef];
+
+    c.bench_function("row_batch 4096 string_dictionary(100, 0.5)", |b| {
+        b.iter(|| {
+            let mut converter = RowConverter::new(vec![SortField::new(DataType::Utf8)]);
+            black_box(converter.convert_columns(&cols))
+        });
+    });
+
     let cols = [
         Arc::new(create_string_array_with_len::<i32>(4096, 0.5, 20)) as ArrayRef,
         Arc::new(create_string_array_with_len::<i32>(4096, 0., 30)) as ArrayRef,
@@ -108,6 +149,30 @@ fn row_bench(c: &mut Criterion) {
             });
         },
     );
+
+    let cols = [
+        Arc::new(create_string_dict_array::<Int32Type>(4096, 0.5, 20)) as ArrayRef,
+        Arc::new(create_string_dict_array::<Int32Type>(4096, 0., 30)) as ArrayRef,
+        Arc::new(create_string_dict_array::<Int32Type>(4096, 0., 100)) as ArrayRef,
+        Arc::new(create_primitive_array::<Int64Type>(4096, 0.)) as ArrayRef,
+    ];
+
+    let fields = [
+        SortField::new(DataType::Utf8),
+        SortField::new(DataType::Utf8),
+        SortField::new(DataType::Utf8),
+        SortField::new(DataType::Int64),
+    ];
+
+    c.bench_function(
+        "row_batch 4096 string_dictionary(20, 0.5), string_dictionary(30, 0), string_dictionary(100, 0), i64(0)",
+        |b| {
+            b.iter(|| {
+                let mut converter = RowConverter::new(fields.to_vec());
+                black_box(converter.convert_columns(&cols))
+            });
+        },
+    );
 }
 
 criterion_group!(benches, row_bench);