71bb0e72ce
This commit is a followup to https://github.com/rust-lang/rust/pull/124032. It replaces the tests that test the various sort functions in the standard library with a test-suite developed as part of https://github.com/Voultapher/sort-research-rs. The current tests suffer a couple of problems: - They don't cover important real world patterns that the implementations take advantage of and execute special code for. - The input lengths tested miss out on code paths. For example, important safety property tests never reach the quicksort part of the implementation. - The miri side is often limited to `len <= 20` which means it very thoroughly tests the insertion sort, which accounts for 19 out of 1.5k LoC. - They are split into to core and alloc, causing code duplication and uneven coverage. - The randomness is not repeatable, as it relies on `std:#️⃣:RandomState::new().build_hasher()`. Most of these issues existed before https://github.com/rust-lang/rust/pull/124032, but they are intensified by it. One thing that is new and requires additional testing, is that the new sort implementations specialize based on type properties. For example `Freeze` and non `Freeze` execute different code paths. Effectively there are three dimensions that matter: - Input type - Input length - Input pattern The ported test-suite tests various properties along all three dimensions, greatly improving test coverage. It side-steps the miri issue by preferring sampled approaches. For example the test that checks if after a panic the set of elements is still the original one, doesn't do so for every single possible panic opportunity but rather it picks one at random, and performs this test across a range of input length, which varies the panic point across them. This allows regular execution to easily test inputs of length 10k, and miri execution up to 100 which covers significantly more code. The randomness used is tied to a fixed - but random per process execution - seed. This allows for fully repeatable tests and fuzzer like exploration across multiple runs. Structure wise, the tests are previously found in the core integration tests for `sort_unstable` and alloc unit tests for `sort`. The new test-suite was developed to be a purely black-box approach, which makes integration testing the better place, because it can't accidentally rely on internal access. Because unwinding support is required the tests can't be in core, even if the implementation is, so they are now part of the alloc integration tests. Are there architectures that can only build and test core and not alloc? If so, do such platforms require sort testing? For what it's worth the current implementation state passes miri `--target mips64-unknown-linux-gnuabi64` which is big endian. The test-suite also contains tests for properties that were and are given by the current and previous implementations, and likely relied upon by users but weren't tested. For example `self_cmp` tests that the two parameters `a` and `b` passed into the comparison function are never references to the same object, which if the user is sorting for example a `&mut [Mutex<i32>]` could lead to a deadlock. Instead of using the hashed caller location as rand seed, it uses seconds since unix epoch / 10, which given timestamps in the CI should be reasonably easy to reproduce, but also allows fuzzer like space exploration.
209 lines
8.0 KiB
Rust
209 lines
8.0 KiB
Rust
// This module implements a Zipfian distribution generator.
|
|
//
|
|
// Based on https://github.com/jonhoo/rust-zipf.
|
|
|
|
use rand::Rng;
|
|
|
|
/// Random number generator that generates Zipf-distributed random numbers using rejection
|
|
/// inversion.
|
|
#[derive(Clone, Copy)]
|
|
pub struct ZipfDistribution {
|
|
/// Number of elements
|
|
num_elements: f64,
|
|
/// Exponent parameter of the distribution
|
|
exponent: f64,
|
|
/// `hIntegral(1.5) - 1}`
|
|
h_integral_x1: f64,
|
|
/// `hIntegral(num_elements + 0.5)}`
|
|
h_integral_num_elements: f64,
|
|
/// `2 - hIntegralInverse(hIntegral(2.5) - h(2)}`
|
|
s: f64,
|
|
}
|
|
|
|
impl ZipfDistribution {
|
|
/// Creates a new [Zipf-distributed](https://en.wikipedia.org/wiki/Zipf's_law)
|
|
/// random number generator.
|
|
///
|
|
/// Note that both the number of elements and the exponent must be greater than 0.
|
|
pub fn new(num_elements: usize, exponent: f64) -> Result<Self, ()> {
|
|
if num_elements == 0 {
|
|
return Err(());
|
|
}
|
|
if exponent <= 0f64 {
|
|
return Err(());
|
|
}
|
|
|
|
let z = ZipfDistribution {
|
|
num_elements: num_elements as f64,
|
|
exponent,
|
|
h_integral_x1: ZipfDistribution::h_integral(1.5, exponent) - 1f64,
|
|
h_integral_num_elements: ZipfDistribution::h_integral(
|
|
num_elements as f64 + 0.5,
|
|
exponent,
|
|
),
|
|
s: 2f64
|
|
- ZipfDistribution::h_integral_inv(
|
|
ZipfDistribution::h_integral(2.5, exponent)
|
|
- ZipfDistribution::h(2f64, exponent),
|
|
exponent,
|
|
),
|
|
};
|
|
|
|
// populate cache
|
|
|
|
Ok(z)
|
|
}
|
|
}
|
|
|
|
impl ZipfDistribution {
|
|
fn next<R: Rng + ?Sized>(&self, rng: &mut R) -> usize {
|
|
// The paper describes an algorithm for exponents larger than 1 (Algorithm ZRI).
|
|
//
|
|
// The original method uses
|
|
// H(x) = (v + x)^(1 - q) / (1 - q)
|
|
// as the integral of the hat function.
|
|
//
|
|
// This function is undefined for q = 1, which is the reason for the limitation of the
|
|
// exponent.
|
|
//
|
|
// If instead the integral function
|
|
// H(x) = ((v + x)^(1 - q) - 1) / (1 - q)
|
|
// is used, for which a meaningful limit exists for q = 1, the method works for all
|
|
// positive exponents.
|
|
//
|
|
// The following implementation uses v = 0 and generates integral number in the range [1,
|
|
// num_elements]. This is different to the original method where v is defined to
|
|
// be positive and numbers are taken from [0, i_max]. This explains why the implementation
|
|
// looks slightly different.
|
|
|
|
let hnum = self.h_integral_num_elements;
|
|
|
|
loop {
|
|
use std::cmp;
|
|
let u: f64 = hnum + rng.gen::<f64>() * (self.h_integral_x1 - hnum);
|
|
// u is uniformly distributed in (h_integral_x1, h_integral_num_elements]
|
|
|
|
let x: f64 = ZipfDistribution::h_integral_inv(u, self.exponent);
|
|
|
|
// Limit k to the range [1, num_elements] if it would be outside
|
|
// due to numerical inaccuracies.
|
|
let k64 = x.max(1.0).min(self.num_elements);
|
|
// float -> integer rounds towards zero, so we add 0.5
|
|
// to prevent bias towards k == 1
|
|
let k = cmp::max(1, (k64 + 0.5) as usize);
|
|
|
|
// Here, the distribution of k is given by:
|
|
//
|
|
// P(k = 1) = C * (hIntegral(1.5) - h_integral_x1) = C
|
|
// P(k = m) = C * (hIntegral(m + 1/2) - hIntegral(m - 1/2)) for m >= 2
|
|
//
|
|
// where C = 1 / (h_integral_num_elements - h_integral_x1)
|
|
if k64 - x <= self.s
|
|
|| u >= ZipfDistribution::h_integral(k64 + 0.5, self.exponent)
|
|
- ZipfDistribution::h(k64, self.exponent)
|
|
{
|
|
// Case k = 1:
|
|
//
|
|
// The right inequality is always true, because replacing k by 1 gives
|
|
// u >= hIntegral(1.5) - h(1) = h_integral_x1 and u is taken from
|
|
// (h_integral_x1, h_integral_num_elements].
|
|
//
|
|
// Therefore, the acceptance rate for k = 1 is P(accepted | k = 1) = 1
|
|
// and the probability that 1 is returned as random value is
|
|
// P(k = 1 and accepted) = P(accepted | k = 1) * P(k = 1) = C = C / 1^exponent
|
|
//
|
|
// Case k >= 2:
|
|
//
|
|
// The left inequality (k - x <= s) is just a short cut
|
|
// to avoid the more expensive evaluation of the right inequality
|
|
// (u >= hIntegral(k + 0.5) - h(k)) in many cases.
|
|
//
|
|
// If the left inequality is true, the right inequality is also true:
|
|
// Theorem 2 in the paper is valid for all positive exponents, because
|
|
// the requirements h'(x) = -exponent/x^(exponent + 1) < 0 and
|
|
// (-1/hInverse'(x))'' = (1+1/exponent) * x^(1/exponent-1) >= 0
|
|
// are both fulfilled.
|
|
// Therefore, f(x) = x - hIntegralInverse(hIntegral(x + 0.5) - h(x))
|
|
// is a non-decreasing function. If k - x <= s holds,
|
|
// k - x <= s + f(k) - f(2) is obviously also true which is equivalent to
|
|
// -x <= -hIntegralInverse(hIntegral(k + 0.5) - h(k)),
|
|
// -hIntegralInverse(u) <= -hIntegralInverse(hIntegral(k + 0.5) - h(k)),
|
|
// and finally u >= hIntegral(k + 0.5) - h(k).
|
|
//
|
|
// Hence, the right inequality determines the acceptance rate:
|
|
// P(accepted | k = m) = h(m) / (hIntegrated(m+1/2) - hIntegrated(m-1/2))
|
|
// The probability that m is returned is given by
|
|
// P(k = m and accepted) = P(accepted | k = m) * P(k = m)
|
|
// = C * h(m) = C / m^exponent.
|
|
//
|
|
// In both cases the probabilities are proportional to the probability mass
|
|
// function of the Zipf distribution.
|
|
|
|
return k;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
impl rand::distributions::Distribution<usize> for ZipfDistribution {
|
|
fn sample<R: Rng + ?Sized>(&self, rng: &mut R) -> usize {
|
|
self.next(rng)
|
|
}
|
|
}
|
|
|
|
use std::fmt;
|
|
impl fmt::Debug for ZipfDistribution {
|
|
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> Result<(), fmt::Error> {
|
|
f.debug_struct("ZipfDistribution")
|
|
.field("e", &self.exponent)
|
|
.field("n", &self.num_elements)
|
|
.finish()
|
|
}
|
|
}
|
|
|
|
impl ZipfDistribution {
|
|
/// Computes `H(x)`, defined as
|
|
///
|
|
/// - `(x^(1 - exponent) - 1) / (1 - exponent)`, if `exponent != 1`
|
|
/// - `log(x)`, if `exponent == 1`
|
|
///
|
|
/// `H(x)` is an integral function of `h(x)`, the derivative of `H(x)` is `h(x)`.
|
|
fn h_integral(x: f64, exponent: f64) -> f64 {
|
|
let log_x = x.ln();
|
|
helper2((1f64 - exponent) * log_x) * log_x
|
|
}
|
|
|
|
/// Computes `h(x) = 1 / x^exponent`
|
|
fn h(x: f64, exponent: f64) -> f64 {
|
|
(-exponent * x.ln()).exp()
|
|
}
|
|
|
|
/// The inverse function of `H(x)`.
|
|
/// Returns the `y` for which `H(y) = x`.
|
|
fn h_integral_inv(x: f64, exponent: f64) -> f64 {
|
|
let mut t: f64 = x * (1f64 - exponent);
|
|
if t < -1f64 {
|
|
// Limit value to the range [-1, +inf).
|
|
// t could be smaller than -1 in some rare cases due to numerical errors.
|
|
t = -1f64;
|
|
}
|
|
(helper1(t) * x).exp()
|
|
}
|
|
}
|
|
|
|
/// Helper function that calculates `log(1 + x) / x`.
|
|
/// A Taylor series expansion is used, if x is close to 0.
|
|
fn helper1(x: f64) -> f64 {
|
|
if x.abs() > 1e-8 { x.ln_1p() / x } else { 1f64 - x * (0.5 - x * (1.0 / 3.0 - 0.25 * x)) }
|
|
}
|
|
|
|
/// Helper function to calculate `(exp(x) - 1) / x`.
|
|
/// A Taylor series expansion is used, if x is close to 0.
|
|
fn helper2(x: f64) -> f64 {
|
|
if x.abs() > 1e-8 {
|
|
x.exp_m1() / x
|
|
} else {
|
|
1f64 + x * 0.5 * (1f64 + x * 1.0 / 3.0 * (1f64 + 0.25 * x))
|
|
}
|
|
}
|