NVIDIA/nvbench
1
0
Fork 0
mirror of https://github.com/NVIDIA/nvbench.git synced 2026-03-14 20:27:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

Compute mean once

Browse Source
This commit is contained in:
Georgy Evtushenko
2024-01-09 14:57:29 -08:00
parent 2dd4c18e2f
commit d079a49976
2 changed files with 46 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
nvbench/detail/entropy_criterion.cxx
View File

@@ -126,16 +126,17 @@ bool entropy_criterion::is_finished()
}
auto begin = m_entropy_tracker.cbegin();
auto end = m_entropy_tracker.cend();
auto end = m_entropy_tracker.cend();
auto mean = statistics::compute_mean(begin, end);
const auto [slope, intercept] = statistics::compute_linear_regression(begin, end);
const auto [slope, intercept] = statistics::compute_linear_regression(begin, end, mean);
if (statistics::slope2deg(slope) > m_max_angle)
{
return false;
}
const auto r2 = statistics::compute_r2(begin, end, slope, intercept);
const auto r2 = statistics::compute_r2(begin, end, mean, slope, intercept);
if (r2 < m_min_r2)
{
return false;

51
nvbench/detail/statistics.cuh
View File

@@ -62,13 +62,22 @@ ValueType standard_deviation(Iter first, Iter last, ValueType mean)
return std::sqrt(variance);
}
template <class It>
nvbench::float64_t compute_mean(It first, It last)
{
const auto n = std::distance(first, last);
return std::accumulate(first, last, 0.0) / static_cast<nvbench::float64_t>(n);
}
/**
* Computes linear regression and returns the slope and intercept
*
* This version takes precomputed mean of [first, last).
* If the input has fewer than 2 samples, infinity is returned for both slope and intercept.
*/
template <class It>
std::pair<nvbench::float64_t, nvbench::float64_t> compute_linear_regression(It first, It last)
std::pair<nvbench::float64_t, nvbench::float64_t>
compute_linear_regression(It first, It last, nvbench::float64_t mean_y)
{
const std::size_t n = static_cast<std::size_t>(std::distance(first, last));
@@ -80,10 +89,9 @@ std::pair<nvbench::float64_t, nvbench::float64_t> compute_linear_regression(It f
// Assuming x starts from 0
const nvbench::float64_t mean_x = (static_cast<nvbench::float64_t>(n) - 1.0) / 2.0;
const nvbench::float64_t mean_y = std::accumulate(first, last, 0.0) / static_cast<nvbench::float64_t>(n);
// Calculate the numerator and denominator for the slope
nvbench::float64_t numerator = 0.0;
nvbench::float64_t numerator = 0.0;
nvbench::float64_t denominator = 0.0;
for (std::size_t i = 0; i < n; ++i, ++first)
@@ -94,22 +102,37 @@ std::pair<nvbench::float64_t, nvbench::float64_t> compute_linear_regression(It f
}
// Calculate the slope and intercept
const nvbench::float64_t slope = numerator / denominator;
const nvbench::float64_t slope = numerator / denominator;
const nvbench::float64_t intercept = mean_y - slope * mean_x;
return std::make_pair(slope, intercept);
}
/**
* Computes and returns the R^2 (coefficient of determination)
*/
* Computes linear regression and returns the slope and intercept
*
* If the input has fewer than 2 samples, infinity is returned for both slope and intercept.
*/
template <class It>
nvbench::float64_t compute_r2(It first, It last, nvbench::float64_t slope, nvbench::float64_t intercept)
std::pair<nvbench::float64_t, nvbench::float64_t> compute_linear_regression(It first, It last)
{
return compute_linear_regression(first, last, compute_mean(first, last));
}
/**
* Computes and returns the R^2 (coefficient of determination)
*
* This version takes precomputed mean of [first, last).
*/
template <class It>
nvbench::float64_t compute_r2(It first,
It last,
nvbench::float64_t mean_y,
nvbench::float64_t slope,
nvbench::float64_t intercept)
{
const std::size_t n = static_cast<std::size_t>(std::distance(first, last));
const nvbench::float64_t mean_y = std::accumulate(first, last, 0.0) / static_cast<nvbench::float64_t>(n);
nvbench::float64_t ss_tot = 0.0;
nvbench::float64_t ss_res = 0.0;
@@ -130,6 +153,16 @@ nvbench::float64_t compute_r2(It first, It last, nvbench::float64_t slope, nvben
return 1.0 - ss_res / ss_tot;
}
/**
* Computes and returns the R^2 (coefficient of determination)
*/
template <class It>
nvbench::float64_t
compute_r2(It first, It last, nvbench::float64_t slope, nvbench::float64_t intercept)
{
return compute_r2(first, last, compute_mean(first, last), slope, intercept);
}
inline nvbench::float64_t rad2deg(nvbench::float64_t rad)
{
return rad * 180.0 / M_PI;