waylonflinn/weblas

Unit Test Float Encode Exhaustively

Open

#11 建立於 2015年12月16日

在 GitHub 查看
 (2 留言) (0 反應) (0 負責人)JavaScript (701 star) (56 fork)batch import
correctnesshelp wanted

描述

The floating point encode step is a tricky bit to get right. Issue #10 shows that numerical stability issues can make a big difference in some edge cases.

Encoding each unique floating point value as an element of a 4098 square matrix should require only a few hundred of these matrices. This should be doable in a reasonable amount of time.

References

貢獻者指南

技術棧
javascript
領域
testing
議題類型
test
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
4
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
1-2 days
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
clear
前置要求
IEEE 754 floating pointWebGL fundamentalsweblas codebase structureunit testing practices
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
20
研究方向
First, review the existing float encode implementation in the weblas source code (likely in a WebGL shader or JavaScript helper). Understand the matrix format (4096 elements per matrix) and the encoding scheme. Then, generate a comprehensive set of test cases covering all IEEE 754 special values (NaN, Infinity, denormals) and edge cases. Use JavaScript's Float32Array to create known floating point values and compare the encoded output against a reference implementation. Check issue #10 for previously identified numerical stability issues that need attention.