frbus_shock: stacking warning only on monetary levers (trp_aerr periods=4 gives +2.37pp, not +1pp); poil_aerr labelled 'model units'

Aberta
  • Algo errado?
#212 1 comentário 0 reações 0 responsáveis Ver no GitHub

Mantenedores costumam responder em até 5 dias

Ninguém assumiu esta issue ainda.

Avaliação

Dificuldade
4/5
Tempo estimado
3-5 dias
Facilidade para iniciantes
48/100
Tipo de issue
Bug
Clareza
Razoavelmente clara
Status de atividade
Ativa
Stack de tecnologia
python
Domínio
api

Direção de pesquisa

Start with integration/src/policyengine_macro/core.py, especially FRBUS_VARIABLES and the fallback near line 730; compare the lever unit strings with the period-stacking guidance in us-frb-model/VALIDATION.md. Check the hosted calls cited in the issue against the documented behavior, then update the affected lever descriptions and poil_aerr units. Done means the exposed levers accurately describe stacking and oil-price units; the optional exogenize mode is a broader follow-up.

Escrita pelo modelo de indexação a partir do texto da issue.

Descrição

bug

What

frbus_shock warns that add-factors stack over periods only for the two monetary levers. Every other lever's units string is silent, and poil_aerr returns "units": "model units". us-frb-model/VALIDATION.md ("Period stacking") says: "Any wrapper exposing a periods argument must document this on every lever it exposes."

Evidence (hosted, us-frb-model@2508a2d, 6 Nov 2026)

Call Label returned What happened
frbus_shock(var="trp_aerr", shock=0.01, periods=4) "0.01 = a 1 percentage point rise in the effective personal tax rate" trp +6.6% → +15.5% of baseline by q4, i.e. +2.37pp, not +1pp. GDP −0.39% vs −0.115% for periods=1
frbus_shock(var="egfe_aerr", shock=0.01, periods=4) (per VALIDATION.md) "0.01 ~ a 1% higher level" level +2.90% by q4, not +1%
frbus_shock(var="poil_aerr", shock=10, periods=4) "model units" the units are $/bbl: +10 gives poil +15.4%. It does not stack (level equation)

Also: the validated sustained +1%-of-GDP purchases experiment (egfe exogenised, VALIDATION.md Test 4, yr-1 multiplier 0.725) cannot be reached through the hosted tool. The nearest call is a one-quarter impulse.

Fix

In integration/src/policyengine_macro/core.py (FRBUS_VARIABLES, ~l.392–450, fallback at l.730):

  1. Add the stacking sentence to egfe_aerr, egse_aerr, trp_aerr, trci_aerr, ecnia_aerr (and any other _aerr lever whose equation carries persistence). Better still, return the realised lever path (variables=[<lever>] by default), so callers see what they got.
  2. Give poil_aerr a real unit ("$/barrel add-error on the oil price level; +10 ≈ +15% at the 2026 LONGBASE price").
  3. Optional: an exogenize mode for egfe/trp, so the Test 4 experiment is reachable hosted.

🤖 Generated with Claude Code

Linguagem predominante
HTML
Estrelas
3
Forks
0
Merge médio
4d 6h
PRs com merge (30d)
15

Preparar o ambiente

Primeiros passos

  1. Leia a issue inteira e depois o guia de contribuição do projeto.
  2. Comente na issue dizendo que vai assumir — evita que duas pessoas façam o mesmo trabalho.
  3. Faça um fork do repositório e trabalhe em uma branch.
  4. Abra um pull request que referencie o número da issue.

Mais de PolicyEngine/macro

Todas as issues de PolicyEngine/macro

Issues semelhantes

Mais issues de Backend & API Design

Receba novas issues na sua caixa de entrada

Um resumo curto de issues do GitHub para quem está começando.