-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Jan-Willem Maessen
committed
Dec 23, 2021
0 parents
commit a45a0b8
Showing
4 changed files
with
376 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,9 @@ | ||
[package] | ||
name = "bkk_hash" | ||
version = "0.1.0" | ||
edition = "2021" | ||
|
||
# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html | ||
|
||
[dependencies] | ||
rand = "*" |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,256 @@ | ||
use std::{cmp::min, collections::hash_map::HashMap, mem::size_of}; | ||
|
||
const SHIFT: usize = 10; | ||
pub const N: usize = 1 << SHIFT; | ||
const RSHIFT: usize = size_of::<usize>() * 8 - SHIFT; | ||
|
||
pub type Key = usize; | ||
|
||
pub type ProbeHist = HashMap<Key, usize>; | ||
|
||
#[derive(Copy, Clone, Debug)] | ||
enum Entry { | ||
Empty, | ||
Full(usize, Key), | ||
Tombstone(usize, Key), | ||
} | ||
|
||
#[derive(Debug)] | ||
pub struct HashSet { | ||
bkk: bool, | ||
size: usize, | ||
pub sum_probe_len: usize, | ||
probe_hist: ProbeHist, | ||
pub probe_hists: Vec<ProbeHist>, | ||
pub insert_lens: Vec<usize>, | ||
set: [Entry; N], | ||
pub set_order: Vec<Key>, | ||
next_clean_size: usize, | ||
curr_x: usize, | ||
} | ||
|
||
pub fn bucket(k: Key) -> usize { | ||
k >> RSHIFT | ||
} | ||
|
||
fn entry_later(i: usize, bk: usize, k: Key, bn: usize, n: Key) -> bool { | ||
if bn <= i { | ||
// No wrap. | ||
if bk <= i { | ||
// No wrap. | ||
n <= k | ||
} else { | ||
false | ||
} | ||
} else { | ||
// wrap | ||
if bk <= i { | ||
// No wrap. | ||
true | ||
} else { | ||
n <= k | ||
} | ||
} | ||
} | ||
|
||
impl HashSet { | ||
pub fn new(bkk: bool) -> Self { | ||
HashSet { | ||
bkk, | ||
size: 0, | ||
sum_probe_len: 0, | ||
probe_hist: ProbeHist::default(), | ||
probe_hists: Vec::with_capacity(N), | ||
insert_lens: Vec::with_capacity(N), | ||
set: [Entry::Empty; N], | ||
set_order: Vec::with_capacity(N + 1), | ||
next_clean_size: N / 4, | ||
curr_x: 1, | ||
} | ||
} | ||
|
||
pub fn size(&self) -> usize { | ||
self.size | ||
} | ||
|
||
// Insert elt with key n, bumping elements as necessary. Returns final | ||
// bucket position (possibly + N if we wrapped around). | ||
fn insert_loop(&mut self, mut bn: usize, mut n: Key, mut b: usize) -> usize { | ||
loop { | ||
let i = b & (N - 1); | ||
match self.set[i] { | ||
Entry::Empty | Entry::Tombstone(_, _) => { | ||
self.set[i] = Entry::Full(bn, n); | ||
self.register_insert(bn, n, b); | ||
break; | ||
} | ||
Entry::Full(_, k) if n == k => { | ||
break; | ||
} | ||
Entry::Full(bk, k) => { | ||
self.set[i] = Entry::Full(bn, n); | ||
self.register_remove(k, b); | ||
self.register_insert(bn, n, b); | ||
n = k; | ||
bn = bk; | ||
if b >= N + bn { | ||
b -= N; | ||
} else if bn > b { | ||
b += N; | ||
} | ||
} | ||
} | ||
b += 1; | ||
} | ||
return b + 1; | ||
} | ||
|
||
pub fn insert(&mut self, n: Key) { | ||
self.set_order.push(n); | ||
let b = self.probe_loc(n); | ||
let bn = bucket(n); | ||
let b = self.insert_loop(bn, n, b); | ||
if self.bkk && self.size == self.next_clean_size { | ||
self.curr_x = N / (N - self.size()); | ||
let x4 = 4 * self.curr_x; | ||
let x2 = 2 * self.curr_x; | ||
let nx2 = N / x2; | ||
for ii in 1..=nx2 { | ||
let i = ii * x2; | ||
let k = (i << RSHIFT) - 1; | ||
self.insert_tombstone(k); | ||
} | ||
self.next_clean_size += N / x4; | ||
} | ||
self.register_insert_len(b - 1, bn); | ||
} | ||
|
||
pub fn remove_tombstone(&mut self, n: Key) { | ||
let mut i = self.probe_loc(n) & (N - 1); | ||
match self.set[i] { | ||
Entry::Tombstone(_, k) if n == k => { } | ||
_ => return, | ||
} | ||
loop { | ||
let i1 = (i + 1) & (N - 1); | ||
match self.set[i1] { | ||
Entry::Full(bk, _) | Entry::Tombstone(bk, _) if bk != i1 => { | ||
self.set[i] = self.set[i1]; | ||
} | ||
_ => { | ||
self.set[i] = Entry::Empty; | ||
return; | ||
} | ||
} | ||
i = i1; | ||
} | ||
} | ||
|
||
fn insert_tombstone(&mut self, n: Key) { | ||
let b = self.probe_loc(n) & (N - 1); | ||
let i = b & (N - 1); | ||
let bn = bucket(n); | ||
match self.set[i] { | ||
Entry::Empty => { } | ||
Entry::Full(_, k) if k == n => { } | ||
Entry::Full(bk, k) => { | ||
self.register_remove(k, b); | ||
self.set[i] = Entry::Tombstone(bn, n); | ||
self.insert_loop(bk, k, b + 1); | ||
} | ||
Entry::Tombstone(_, _) => { | ||
self.set[i] = Entry::Tombstone(bn, n); | ||
} | ||
} | ||
} | ||
|
||
fn register_remove(&mut self, k: Key, b: usize) { | ||
let bk = bucket(k); | ||
let mut d = if bk > b { b + N - bk } else { b - bk }; | ||
if d >= N { | ||
d -= N; | ||
} | ||
self.sum_probe_len -= d; | ||
*self.probe_hist.get_mut(&d).unwrap() -= 1; | ||
self.size -= 1; | ||
} | ||
|
||
fn register_insert(&mut self, b0: usize, _: Key, b: usize) { | ||
let d = if b0 <= b { | ||
b - b0 | ||
} else { | ||
b + N - b0 | ||
}; | ||
self.sum_probe_len += d; | ||
*self.probe_hist.entry(d).or_insert(0) += 1; | ||
self.size += 1; | ||
} | ||
|
||
fn register_insert_len(&mut self, b: usize, b0: usize) { | ||
let d = if b < b0 { b + N - b0 } else { b - b0 }; | ||
self.insert_lens.push(d); | ||
self.probe_hists.push(self.probe_hist.clone()); | ||
} | ||
|
||
fn probe_loc(&self, n: Key) -> usize { | ||
let bn = bucket(n); | ||
let mut b = bn; | ||
loop { | ||
let i = b & (N - 1); | ||
match self.set[i] { | ||
Entry::Empty => break, | ||
Entry::Tombstone(bk, k) | Entry::Full(bk, k) => { | ||
if n == k || entry_later(i, bk, k, bn, n) { | ||
break; | ||
} | ||
} | ||
} | ||
b += 1; | ||
} | ||
b | ||
} | ||
|
||
pub fn probe_len(&self, n: Key) -> usize { | ||
let bn = bucket(n); | ||
let b = self.probe_loc(n); | ||
b - bn | ||
} | ||
|
||
pub fn iter(&'_ self) -> HashSetIter<'_> { | ||
HashSetIter { | ||
set: self, | ||
next_bucket: 0, | ||
} | ||
} | ||
} | ||
|
||
pub struct HashSetIter<'a> { | ||
set: &'a HashSet, | ||
next_bucket: usize, | ||
} | ||
|
||
impl<'a> Iterator for HashSetIter<'a> { | ||
type Item = &'a Key; | ||
|
||
fn size_hint(&self) -> (usize, Option<usize>) { | ||
let upper = min(N - self.next_bucket, self.set.size); | ||
let lower = if self.next_bucket <= self.set.size { | ||
self.set.size - self.next_bucket | ||
} else { | ||
0 | ||
}; | ||
(lower, Some(upper)) | ||
} | ||
|
||
fn next(&mut self) -> Option<Self::Item> { | ||
while self.next_bucket < N { | ||
let b = self.next_bucket; | ||
self.next_bucket += 1; | ||
match &self.set.set[b] { | ||
Entry::Full(_, k) => return Some(k), | ||
_ => {} | ||
} | ||
} | ||
return None; | ||
} | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,2 @@ | ||
#![allow(dead_code)] | ||
pub mod hash_set; |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,109 @@ | ||
use bkk_hash::hash_set::{bucket, HashSet, Key, ProbeHist, N}; | ||
use rand::random; | ||
|
||
fn dump_hists(desc: &str, hists: Vec<ProbeHist>) { | ||
for (i, h) in hists.into_iter().enumerate() { | ||
let x = N as f64 / (N - i) as f64; | ||
print!("{}, {:3}, {:.3},", desc, i, x); | ||
let mx = h.iter().map(|(&n, _)| n).max().unwrap_or(0); | ||
let sum_x: usize = h.iter().map(|(&n, &v)| n * v).sum(); | ||
let sum_x = sum_x as f64; | ||
let sum_x2: usize = h.iter().map(|(&n, &v)| n * n * v).sum(); | ||
let sum_x2 = sum_x2 as f64; | ||
let n: usize = h.iter().map(|(_, &v)| v).sum(); | ||
let n = n as f64; | ||
let mean = sum_x / n; | ||
let stddev = ((sum_x2 / n) - (mean * mean)).sqrt(); | ||
print!("\t{:3.3},\t{:3.3},", mean, stddev); | ||
print!("\t{:3.3},\t{:3.3},", mean / x, stddev / x); | ||
for l in 0..=mx { | ||
if let Some(&n) = h.get(&l) { | ||
print!("\t{},", n); | ||
} else { | ||
print!("\t,"); | ||
} | ||
} | ||
println!(""); | ||
} | ||
} | ||
|
||
fn agg_to_hists(hists: &mut Vec<ProbeHist>, new_hists: Vec<ProbeHist>) { | ||
for (i, h) in new_hists.into_iter().enumerate() { | ||
let hist = &mut hists[i]; | ||
for (k, v) in h.into_iter() { | ||
*hist.entry(k).or_default() += v; | ||
} | ||
} | ||
} | ||
|
||
fn do_one(bkk: bool) { | ||
let mut agg_probe_hists = vec![ProbeHist::default(); N]; | ||
let mut agg_insert_hists = vec![ProbeHist::default(); N]; | ||
for trial in 0..1000 { | ||
let mut set = HashSet::new(bkk); | ||
while set.size() < N { | ||
let i: Key = random(); | ||
set.insert(i); | ||
let mut h = ProbeHist::new(); | ||
let mut sum_probe_len = 0; | ||
for &j in set.iter() { | ||
let l = set.probe_len(j); | ||
sum_probe_len += l; | ||
*h.entry(l).or_insert(0) += 1; | ||
} | ||
if sum_probe_len != set.sum_probe_len { | ||
for &j in set.iter() { | ||
println!("pl {:3} {} = {:3}", bucket(j), j, set.probe_len(j)); | ||
} | ||
} | ||
assert_eq!(sum_probe_len, set.sum_probe_len, "{:#?}", set); | ||
set.probe_hists.last_mut().unwrap().retain(|_, v| *v != 0); | ||
assert_eq!(&h, set.probe_hists.last().unwrap()); | ||
set.probe_hists.last().unwrap().iter().all(|(k, _)| *k <= set.size()); | ||
assert_eq!( | ||
set.probe_hists | ||
.last() | ||
.unwrap() | ||
.iter() | ||
.map(|(_, v)| *v) | ||
.sum::<usize>(), | ||
set.size() | ||
); | ||
assert_eq!( | ||
set.probe_hists | ||
.last() | ||
.unwrap() | ||
.iter() | ||
.map(|(l, v)| *l * *v) | ||
.sum::<usize>(), | ||
set.sum_probe_len | ||
); | ||
assert_eq!(set.size(), set.insert_lens.len()); | ||
} | ||
if true { | ||
println!( | ||
"{:3},\t{:.3}", | ||
trial, | ||
(set.sum_probe_len as f64 / set.size() as f64), | ||
); | ||
} | ||
if set.probe_hists[N - 1].get(&(N - 1)).unwrap_or(&0) > &0 { | ||
for (i, &v) in set.iter().enumerate() { | ||
println!("{:3}: {:3} {}", i, bucket(v), v); | ||
} | ||
panic!("Full probe len!\n{:#?}", set.set_order); | ||
} | ||
agg_to_hists(&mut agg_probe_hists, set.probe_hists); | ||
for (i, &l) in set.insert_lens.iter().enumerate() { | ||
*agg_insert_hists[i].entry(l).or_insert(0) += 1; | ||
} | ||
} | ||
println!("what, i, x,\tmean,\tstddev,\tmean/x,\tstddev/x"); | ||
dump_hists("pr", agg_probe_hists); | ||
dump_hists("in", agg_insert_hists); | ||
} | ||
|
||
fn main() { | ||
do_one(true); | ||
do_one(false); | ||
} |