pymc-devs/pymc

Implement more Dims distributions

Open

#7,874 创建于 2025年7月25日

在 GitHub 查看
 (1 评论) (0 反应) (0 负责人)Python (7,926 star) (1,902 fork)batch import
dimshelp wanted

描述

Description

We have these currently:

We should implement most of the existing distributions. Univariate should be straightforward.

Multivariate may be trickier. We would likely want to reuse the regular methods like we do for ZeroSumNormal and MvNormal, but we need to test carefully.

Distribution factories like Mixture, Truncated, Censored will be the trickiest. The first step is to implement an extend_dist_dims function (to recreate the input distributions with more dims, just like we used to having to change_dist_size before).

贡献者指南

技术栈
python
领域
datamachine learning
议题类型
feature
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
5
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
over 1 week
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
fresh
清晰度议题是否清楚说明期望改动、验收标准和下一步。
clear
前置要求
Understanding of PyMC dims systemFamiliarity with existing Scalar and Vector distributionsBasic knowledge of probability distributions
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
20
研究方向
The issue outlines a plan to implement more Dims distributions. Start by examining the existing Scalar and Vector distribution implementations in pymc/dims/distributions/scalar.py and vector.py. Understand the pattern. Then implement an `extend dist dims` function as described. For multivariate distributions, refer to how ZeroSumNormal and MvNormal reuse regular methods. For distribution factories like Mixture, Truncated, Censored, study the current change dist size pattern. The issue has no comments or linked PRs yet, so the approach is not further clarified.