florian/search_benchmarks
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

bug fix

Browse Source
This commit is contained in:
Florian Stecker 2024-02-23 10:02:54 -05:00
parent feadef4723
commit fa3f3006f3
2 changed files with 27 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
src/main.rs
View File

@ -5,6 +5,7 @@ fn main() {
// generate 800MB of random data // generate 800MB of random data
let mut data: Vec<u64> = vec![0; 100*1024*1024]; let mut data: Vec<u64> = vec![0; 100*1024*1024];
thread_rng().fill(&mut data[..]); thread_rng().fill(&mut data[..]);
data.sort();
let mut metadata: Vec<(usize, usize)> = Vec::new(); let mut metadata: Vec<(usize, usize)> = Vec::new();
let mut all_times: Vec<Vec<(u128, u128, u128)>> = Vec::new(); let mut all_times: Vec<Vec<(u128, u128, u128)>> = Vec::new();
@ -35,17 +36,17 @@ fn main() {
} }
fn test_searches(data: &[u64], runs: usize, items: usize) -> Vec<Duration> { fn test_searches(data: &[u64], runs: usize, items: usize) -> Vec<Duration> {
let needle_offset = random::<u64>(); let needle_offset = random::<usize>();
// let offset_offset = random::<usize>(); let offset_offset = random::<usize>();
let offset_offset = items; // let offset_offset = items;
flush_caches(); flush_caches();
let start_time = Instant::now(); let start_time = Instant::now();
for i in 0..runs { for i in 0..runs {
let offset = usize::wrapping_mul(i+1, offset_offset) % (data.len() - items); let offset = usize::wrapping_mul(i+1, offset_offset) % (data.len() - items);
let needle = u64::wrapping_mul((i+1) as u64, needle_offset); let needle_index = usize::wrapping_mul(i+1, needle_offset) % items;
search_linear_iter(&data[offset..offset+items], needle); search_linear_iter(&data[offset..offset+items], data[offset+needle_index]);
} }
let linear_iter_duration = Instant::now().duration_since(start_time); let linear_iter_duration = Instant::now().duration_since(start_time);
@ -54,8 +55,8 @@ fn test_searches(data: &[u64], runs: usize, items: usize) -> Vec<Duration> {
let start_time = Instant::now(); let start_time = Instant::now();
for i in 0..runs { for i in 0..runs {
let offset = usize::wrapping_mul(i+1, offset_offset) % (data.len() - items); let offset = usize::wrapping_mul(i+1, offset_offset) % (data.len() - items);
let needle = u64::wrapping_mul((i+1) as u64, needle_offset); let needle_index = usize::wrapping_mul(i+1, needle_offset) % items;
search_linear_loop(&data[offset..offset+items], needle); search_linear_loop(&data[offset..offset+items], data[offset+needle_index]);
} }
let linear_loop_duration = Instant::now().duration_since(start_time); let linear_loop_duration = Instant::now().duration_since(start_time);
@ -64,8 +65,8 @@ fn test_searches(data: &[u64], runs: usize, items: usize) -> Vec<Duration> {
let start_time = Instant::now(); let start_time = Instant::now();
for i in 0..runs { for i in 0..runs {
let offset = usize::wrapping_mul(i+1, offset_offset) % (data.len() - items); let offset = usize::wrapping_mul(i+1, offset_offset) % (data.len() - items);
let needle = u64::wrapping_mul((i+1) as u64, needle_offset); let needle_index = usize::wrapping_mul(i+1, needle_offset) % items;
search_binary(&data[offset..offset+items], needle); search_binary(&data[offset..offset+items], data[offset+needle_index]);
} }
let binary_duration = Instant::now().duration_since(start_time); let binary_duration = Instant::now().duration_since(start_time);
@ -74,6 +75,7 @@ fn test_searches(data: &[u64], runs: usize, items: usize) -> Vec<Duration> {
vec![linear_iter_duration, linear_loop_duration, binary_duration] vec![linear_iter_duration, linear_loop_duration, binary_duration]
} }
#[allow(dead_code)]
fn flush_caches() { fn flush_caches() {
let mut data: Vec<u64> = vec![0; 1024*1024]; let mut data: Vec<u64> = vec![0; 1024*1024];
thread_rng().fill(&mut data[..]); thread_rng().fill(&mut data[..]);
@ -96,9 +98,9 @@ fn search_binary(haystack: &[u64], needle: u64) {
fn search_linear_iter(haystack: &[u64], needle: u64) { fn search_linear_iter(haystack: &[u64], needle: u64) {
let index = haystack.iter() let index = haystack.iter()
.enumerate() .enumerate()
.skip_while(|(_,x)|**x != needle) .skip_while(|&(_,&x)|x < needle)
.next() .next()
.map(|x|x.0) .map(|(i,_)|i)
.unwrap_or_else(||haystack.len()); .unwrap_or_else(||haystack.len());
if index as u64 == needle { if index as u64 == needle {
@ -109,7 +111,7 @@ fn search_linear_iter(haystack: &[u64], needle: u64) {
fn search_linear_loop(haystack: &[u64], needle: u64) { fn search_linear_loop(haystack: &[u64], needle: u64) {
let mut index = 0; let mut index = 0;
for i in 0..haystack.len() { for i in 0..haystack.len() {
if haystack[i] == needle { if haystack[i] >= needle {
index = i; index = i;
break; break;
} }

20
times.plt
View File

@ -5,11 +5,17 @@ set yrange [1:3000]
set grid set grid
set terminal pngcairo size 1024, 1024 set terminal pngcairo size 1024, 1024
set output "times.png" set output "times_d8.png"
plot "data_c8" using 2:($3/$1/4) w l lw 2 t "linear, iterator, consecutive", \ plot "<paste data_d7 data_d8" \
"" using 2:($4/$1/4) w l lw 2 t "linear, loop, consecutive", \ using 2:($3/$1/4) w l lw 2 t "linear, iterator, consecutive", \
"" using 2:($5/$1/4) w l lw 2 t "binary, consecutive", \ "" using 2:($4/$1/4) w l lw 2 t "linear, loop, consecutive", \
"data_c7" using 2:($3/$1/4) w l lw 2 t "linear, iterator, spread", \ "" using 2:($5/$1/4) w l lw 2 t "binary, consecutive", \
"" using 2:($4/$1/4) w l lw 2 t "linear, loop, spread", \ "" using 2:($8/$1/4) w l lw 2 t "linear, iterator, spread", \
"" using 2:($5/$1/4) w l lw 2 t "binary, spread" "" using 2:($9/$1/4) w l lw 2 t "linear, loop, spread", \
"" using 2:($10/$1/4) w l lw 2 t "binary, spread"
# plot "<paste data_d7 data_d8" \
# using 2:(($8-$3)/$1/4) w l lw 2 t "linear, iterator", \
# "" using 2:(($9-$4)/$1/4) w l lw 2 t "linear, loop", \
# "" using 2:(($10-$5)/$1/4) w l lw 2 t "binary"