From c1e5d709323aac02f441d20fa6479710f4a3e82a Mon Sep 17 00:00:00 2001 From: Matt McKay Date: Thu, 23 Jul 2026 15:01:21 +1000 Subject: [PATCH] =?UTF-8?q?Glossary:=20=E8=BE=B9=E9=99=85=E5=88=86?= =?UTF-8?q?=E5=B8=83=20=E2=86=92=20=E8=BE=B9=E7=BC=98=E5=88=86=E5=B8=83=20?= =?UTF-8?q?for=20"marginal=20distribution"?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The AI review on #198 flagged this and the PR merged before it was fixed, so the term is currently on main. The glossary is explicit and context-aware: "Marginal distribution" maps to 边缘分布 tagged `"context": "probability"`, while every economics sense correctly keeps 边际 — 边际产量, 边际消费倾向, 边际收入, 边际效用. This lecture uses the probability sense throughout: ψ_T is the distribution of X_T at a fixed time T, and the later reference is to the sequence {ψ_t} converging to a limiting distribution. 边缘分布 is the standard mathematical term for that; 边际 is "marginal" as in marginal cost. Four occurrences: the H2 heading, its heading-map value, and two body references. The heading-map value is updated alongside the heading deliberately — leaving them out of step would make the next sync mis-map the section. --- lectures/inventory_dynamics.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/lectures/inventory_dynamics.md b/lectures/inventory_dynamics.md index 7ce39b5..8e3f66f 100644 --- a/lectures/inventory_dynamics.md +++ b/lectures/inventory_dynamics.md @@ -14,7 +14,7 @@ translation: headings: Overview: 概述 Sample paths: 样本路径 - Marginal distributions: 边际分布 + Marginal distributions: 边缘分布 Exercises: 练习 --- @@ -191,9 +191,9 @@ for i in range(400): plt.show() ``` -## 边际分布 +## 边缘分布 -现在让我们来看看某一固定时间点 $T$ 时 $X_T$ 的边际分布 $\psi_T$。 +现在让我们来看看某一固定时间点 $T$ 时 $X_T$ 的边缘分布 $\psi_T$。 我们将通过在给定初始条件 $X_0$ 的情况下，生成多个 $X_T$ 的样本来实现。 @@ -302,7 +302,7 @@ plt.show() （作为背景知识，有关平稳性的讨论，请参见 {doc}`我们关于AR(1)过程的讲座 `——基本概念是相同的。） -特别是，边际分布序列$\{\psi_t\}$正在收敛到一个唯一的极限分布，且该分布不依赖于初始条件。 +特别是，边缘分布序列$\{\psi_t\}$正在收敛到一个唯一的极限分布，且该分布不依赖于初始条件。 虽然我们不会在此证明这一点，但我们可以通过模拟来研究这一性质。