EZPPFarm/akatsuki-pp-rs
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

about accuracy & speed in docs

Browse Source
This commit is contained in:
MaxOhn
2024-02-27 11:47:55 +01:00
parent e81c0e890f
commit 50b06478c7
2 changed files with 66 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+33 -14
View File
@@ -6,21 +6,9 @@
Library to calculate difficulty and performance attributes for all [osu!] gamemodes.
### Description
A large part of `rosu-pp` is a port of [osu!lazer]'s difficulty and performance calculation
with emphasis on a precise translation to Rust for the most accurate results.
Another important factor is the calculation speed. Optimizations and an accurate translation
unfortunately don't always go hand-in-hand. Nonetheless, performance improvements are still
snuck in wherever possible, providing a significantly faster runtime than the native C# code.
TODO: values to compare
Additionally, `rosu-pp` allows previous values to be re-used so that they don't need to be
calculated again. For example, a beatmap needs to be decoded only once and can then be used
for any amount of attribute calculations. Similarly, previous attributes can be re-used for
later calculations (with some limitations, see the [example](#usage)).
with emphasis on a precise translation to Rust for the most [accurate results](#accuracy)
while also providing a significant [boost in performance](#speed).
Last commits of the ported code:
- [osu!lazer] : `7342fb7f51b34533a42bffda89c3d6c569cc69ce` (2022-10-11)
@@ -74,6 +62,36 @@ println!("Stars: {stars} | PP: {pp}/{max_pp}");
TODO
### Accuracy
`rosu-pp` was tested against all current beatmaps on multiple mod combinations and delivered
values that matched osu!lazer perfectly down to the last decimal place.
However, there is one small caveat: the values are only this precise on debug mode.
On release mode, Rust's compiler performs optimizations that produce the tiniest discrepancies
due to floating point inaccuracies which can cascade into larger differences in the end.
With this in mind, `rosu-pp` is still as accurate as can be without targeting the
.NET compiler itself. Realistically, the inaccuracies in release mode are negligibly small.
### Speed
An important factor for `rosu-pp` is the calculation speed. Optimizations and an accurate translation
unfortunately don't always go hand-in-hand. Nonetheless, performance improvements are still
snuck in wherever possible, providing a significantly faster runtime than the native C# code.
Results of a rudimentary [benchmark] of osu!lazer and rosu-pp:
```txt
osu!lazer:
Decoding maps: Median: 378.10ms | Mean: 381.47ms
Calculating difficulties: Median: 588.89ms | Mean: 597.11ms
Calculating performances: Median: 315.90µs | Mean: 310.60µs
rosu-pp:
Decoding maps: Median: 46.94ms | Mean: 47.21ms
Calculating difficulties: Median: 72.90ms | Mean: 73.13ms
Calculating performances: Median: 44.13µs | Mean: 45.53µs
```
### Features
| Flag | Description | Dependencies
@@ -94,5 +112,6 @@ Using `rosu-pp` from other languages than Rust:
[`tracing`]: https://docs.rs/tracing
[rosu-pp-js]: https://github.com/MaxOhn/rosu-pp-js
[rosu-pp-py]: https://github.com/MaxOhn/rosu-pp-py
[benchmark]: https://gist.github.com/MaxOhn/625af10011f6d7e13a171b08ccf959ff
<!-- cargo-rdme end -->
src/lib.rs
+33 -14
View File
@@ -1,20 +1,8 @@
//! Library to calculate difficulty and performance attributes for all [osu!] gamemodes.
//!
//! ## Description
//!
//! A large part of `rosu-pp` is a port of [osu!lazer]'s difficulty and performance calculation
//! with emphasis on a precise translation to Rust for the most accurate results.
//!
//! Another important factor is the calculation speed. Optimizations and an accurate translation
//! unfortunately don't always go hand-in-hand. Nonetheless, performance improvements are still
//! snuck in wherever possible, providing a significantly faster runtime than the native C# code.
//!
//! TODO: values to compare
//!
//! Additionally, `rosu-pp` allows previous values to be re-used so that they don't need to be
//! calculated again. For example, a beatmap needs to be decoded only once and can then be used
//! for any amount of attribute calculations. Similarly, previous attributes can be re-used for
//! later calculations (with some limitations, see the [example](#usage)).
//! with emphasis on a precise translation to Rust for the most [accurate results](#accuracy)
//! while also providing a significant [boost in performance](#speed).
//!
//! Last commits of the ported code:
//! - [osu!lazer] : `7342fb7f51b34533a42bffda89c3d6c569cc69ce` (2022-10-11)
@@ -68,6 +56,36 @@
//!
//! TODO
//!
//! ## Accuracy
//!
//! `rosu-pp` was tested against all current beatmaps on multiple mod combinations and delivered
//! values that matched osu!lazer perfectly down to the last decimal place.
//!
//! However, there is one small caveat: the values are only this precise on debug mode.
//! On release mode, Rust's compiler performs optimizations that produce the tiniest discrepancies
//! due to floating point inaccuracies which can cascade into larger differences in the end.
//! With this in mind, `rosu-pp` is still as accurate as can be without targeting the
//! .NET compiler itself. Realistically, the inaccuracies in release mode are negligibly small.
//!
//! ## Speed
//!
//! An important factor for `rosu-pp` is the calculation speed. Optimizations and an accurate translation
//! unfortunately don't always go hand-in-hand. Nonetheless, performance improvements are still
//! snuck in wherever possible, providing a significantly faster runtime than the native C# code.
//!
//! Results of a rudimentary [benchmark] of osu!lazer and rosu-pp:
//! ```txt
//! osu!lazer:
//! Decoding maps: Median: 378.10ms | Mean: 381.47ms
//! Calculating difficulties: Median: 588.89ms | Mean: 597.11ms
//! Calculating performances: Median: 315.90µs | Mean: 310.60µs
//!
//! rosu-pp:
//! Decoding maps: Median: 46.94ms | Mean: 47.21ms
//! Calculating difficulties: Median: 72.90ms | Mean: 73.13ms
//! Calculating performances: Median: 44.13µs | Mean: 45.53µs
//! ```
//!
//! ## Features
//!
//! | Flag | Description | Dependencies
@@ -88,6 +106,7 @@
//! [`tracing`]: https://docs.rs/tracing
//! [rosu-pp-js]: https://github.com/MaxOhn/rosu-pp-js
//! [rosu-pp-py]: https://github.com/MaxOhn/rosu-pp-py
//! [benchmark]: https://gist.github.com/MaxOhn/625af10011f6d7e13a171b08ccf959ff
#![deny(rustdoc::broken_intra_doc_links, rustdoc::missing_crate_level_docs)]
#![warn(clippy::missing_const_for_fn, clippy::pedantic)]