probcomp/bayeslite

latent variables cannot be inputs to CGPM foreign models

Open

#445 建立於 2016年7月12日

在 GitHub 查看
 (0 留言) (0 反應) (0 負責人)Python (915 star) (63 fork)batch import
help wantedk min/hr

描述

create generator g for p using cgpm (
    model foo given err using venturescript (sp = errfoo),
    latent err numerical
)

This will not work because cgpm_metamodel._create_schema assumes that every input to a foreign model exists as a manifest variable of the population, and passes them to core.bayesdb_variable_stattype to determine the statistical type.

The logic in _create_schema must be changed to defer whatever processing requires the statistical types of inputs to foreign models until after all clauses have been ingested, and then update that processing to use latents (the mapping from latent variable names to their respective statistical types) instead of core.bayesdb_variable_stattype when appropriate.

貢獻者指南

技術棧
python
領域
backenddata
議題類型
bug
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。
half day
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。
mostly clear
前置要求
Pythonbayeslite internalsCGPM conceptbasic Bayesian statistics
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。
10
研究方向
The issue is in `cgpm metamodel. create schema` in the bayeslite codebase. It currently assumes all inputs to foreign models are manifest variables, but latent variables should also be allowed. The fix requires deferring the statistical type determination for foreign model inputs until after all clauses are processed, then using the `latents` mapping instead of `core.bayesdb variable stattype`. Start by examining the ` create schema` method and understanding how it processes foreign models, then design a change that separates the input validation from type resolution.