paul-buerkner/brms

Consider replacing `asym_laplace_*` with native version from Stan?

Open

#1,836 建立於 2025年11月13日

在 GitHub 查看
 (2 留言) (0 反應) (0 負責人)R (1,402 star) (220 fork)batch import
Stanefficiencygood first issue

描述

Stan since 2.28 has natively adopted Asym Laplace and named skew double exponential distribution.

Perhaps it would speed up the sampler by using the native version for asym_laplace family?

貢獻者指南

技術棧
r
領域
datamachine learning
議題類型
performance
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。
3-5 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。
fresh
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
needs investigation
前置要求
R programmingStan knowledgebrms package structure
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。
25
研究方向
Investigate the current implementation of `asym laplace *` functions in the brms source code (likely in `R/` or `src/`). Compare with the native Asymmetric Laplace distribution added in Stan 2.28. Check the brms documentation and any related issues or PRs. The goal is to replace the custom implementation with a call to the native Stan function, potentially improving sampler speed. Further clarification from maintainers on desired approach and compatibility requirements is needed.