Japanese

At a glance

Edition

QuantEcon/lecture-python-programming.ja, translated from Python Programming for Economics and Finance. The edition's website is not yet published.

Phase

Initial translation, starting. Each lecture is drafted and reviewed on its own pull request, one at a time.

Translators

Chihiro Watanabe (@Chihiro2000GitHub) and Kenko Li (@xuanguang-li).

Rulings and merges

The edition has no editor of record. @mmcky rules on terms and house style after consulting both translators, and merges every pull request.

Who reviews which lectures

The book is split by part. Chihiro reviews the opening page and the parts Introduction to Python and Foundations of Scientific Computing (13 pages). Kenko reviews High Performance Computing, Working with Data, More Python Programming and Other (14 pages).

Your review issues

Each lecture has a review issue assigned to its translator, such as “Review: python_by_example.md”. See Chihiro's issues and Kenko's issues. The merge of a lecture's pull request closes its issue.

Lecture order

The edition's REVIEWING.md lists the lectures of each half in order. It follows this manual, and points here for terms and house style.

Rounds in this edition

Each round follows How a review round works. Two points apply to this edition only.

• Chihiro's first page, intro.md, is a short warm-up. It arrives on the pull request that creates the edition's website. That pull request also gives the book's six part titles in Japanese, and both translators can suggest changes to them there.
• Questions about a term go to one follow-up issue for the round, assigned to both translators, because a term chosen in one half applies to the whole book. REVIEWING.md calls this issue a Decision issue.

Term policy

Use Japanese only where a well-known, standard Japanese term exists. Otherwise use English, in Latin script. If in doubt, use English.

• Japanese, because a standard term exists: 予算制約 (budget constraint), 行列 (matrix), 均衡 (equilibrium).
• English, because no standard term exists: forcing variable, forward-looking difference equation.
• Mixed, where only part of a term has a standard Japanese form: lake モデル, cost-to-go 関数.

Programming words, such as function, list and namespace, are not settled yet. A separate glossary pull request will settle them, with numbered questions for both translators. That work is tracked on action-translation#339.

Names and abbreviations

• A person's name stays in Latin script, with no katakana. This includes Japanese and Chinese researchers.
• A term named after people follows the glossary, with the names joined by ・ and never by ＝: ソロー・スワン成長モデル.
• An abbreviation is spelt out on first use, followed by the abbreviation in full-width parentheses: 国内総生産（GDP）.
• FRED and NBER stay as the bare abbreviation, as the English writes them.

Glossary

The engine follows the Japanese glossary, glossary/ja.json in the action-translation repository. The file goes live when action-translation#69 merges. Until then, the link does not work, and the glossary is on that pull request.

The glossary on that pull request was written before the rulings of 2026-10-02. Where it differs from the settled terms below, the settled terms apply.

If a glossary term reads wrong in a lecture, say so in your review. See Terms and house style.

House style

The house style is ruled on action-translation#337. Each ruling links to the comment where it was made.

Settled terms

• A named dataset or survey keeps its English name, unless the glossary gives a Japanese one (2026-10-02, T3).
• 市場清算 stays because 市場均衡 would clash with 均衡, the glossary's term for equilibrium.
• 産業連関モデル stays because 産業連関表 means the input-output table itself.

Open questions

These questions are open on action-translation#337, and the answers shape every lecture drafted after them. Answer on the issue, by question number. A partial answer helps too.

• H1, register. です・ます throughout, である throughout, or です・ます for explanation and である for definitions. See the question.
• H2, sentence punctuation. Choose between 、。, ，． and ，。 as the comma and full stop. See the question.
• H5, code comments. Translate the comments in code cells into Japanese, or keep them in English. See the question.
• C1, katakana glosses. 通貨発行益（シニョリッジ） and 台（サポート） keep a katakana gloss, while 年金 (annuity) has none. Keep both glosses, drop both, or decide each. See the question.

What to skip

Skip what every reviewer skips: see What to look for. In this edition, also skip:

• figure labels, which stay in English for now;
• text inside code, such as comments, unless the draft translates it.

Sources

• Japanese Wikipedia, for terms in economics, mathematics and statistics. For example, T2 follows the article 限界生産力.
• The Japanese Python documentation, for programming words. It is only partly translated. Its glossary gives 名前空間 for namespace, but entries such as mutable and iterator are still in English. Where the documentation is silent, the term policy decides.

Rendering

Figure labels stay in English until the engine supports a Japanese font. Japanese text drawn in a figure without a Japanese font shows as empty boxes.

If Japanese text in a preview shows as empty boxes, or in Chinese character forms, say so on the pull request.

Rulings log

Newest first. Each entry links to where the ruling was made.