Repository navigation
Conversation
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A mutation that opened the leaver routes to a student who fails the entry condition (reg. 5(3A)) passed every test: the grid fixed the entry age at 17. The grid now also draws a course begun at 19, with the date fact set directly, and asserts that such a student never qualifies. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A mutation that read a 19th birthday on 1 September as its own "1 September following" was caught by the calendar property test only. Case 11 pins the edge in YAML too. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…o ctc-qyp-school-leaver-routes Brings in #2285's review round 2 head through the item 1 branch. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…to qyp-income-support-exclusion Brings in the item 2 grid fix and #2285's review round 2 head. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ncome-support-exclusion Brings in the 1 September YAML case. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…e education Review round 1 of #2301: the formula took failing the non-advanced education test as having left education, so a 16-year-old in full-time advanced education working 24 hours a week lost the reg. 4 period. Reg. 5(4)(a) bars work only for a person "having ceased to receive full-time education or approved training", and full-time education at a school or college counts at any level (reg. 5(5)). Reg. 4(2) lifts the first condition only. The 20 weeks of reg. 5(3)(b) are for a person who has ceased. The property oracle had the same conflation and is fixed with it; the grid now draws a course begun at 19, as its comment says. YAML Cases 15 and 16 pin the advanced student. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…r 20 Review round 1 of #2300: setting the 19th-birthday fact for one person gave everyone else in the situation false (the default), so another 19-year-old student silently lost the Pension Credit and Universal Credit status. The formula gives true for everyone under 20 on 6 October, which is everyone the age limit lets be a qualifying young person, so true is now the default for the fact and for the two terminal-date variables that read it. A partial setting then leaves everyone else where the formula puts them. Tested from age alone for ages 0 to 25 and all months, and in a mixed household (property test and YAML Case 12). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review round 1 of #2302: receives_income_support_in_own_right stayed true under a reform that neutralizes income_support, so a young person's report still excluded them from the Pension Credit, Child Benefit and CTC qualifying young person status. It now also needs Income Support not to be neutralized, as is_on_income_support already does for another member's own claim. New test: the same student is excluded with Income Support and qualifies for all three once a reform removes it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Case 3's parent, not set, now takes the default (true): harmless, since every reader is behind the qualifying young person age limit. The older child_qualifying_young_person.yaml case left its after-terminal-date person unset and relied on the old false default; it now sets the input explicitly. The docstrings now say the default is the formula's value under 20 and that older people are excluded before it is read. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ent's means tests Review round 1 of #2299: a young person who qualifies through reg. 7 and works 24 hours stays a young person, so their earnings must not enter the parent's Housing Benefit, Income Support or Council Tax Reduction income. #2387, merged in here, makes that so; Case 11 now has £20,000 of earnings and asserts the three applicable incomes are nil. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Brings in main e01f2f5. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…education Review round 2 of #2301: the formula took apprenticeship as ceasing education, so a 16-year-old apprentice still on a college course and working 24 hours or more lost the reg. 4 period. Education received by virtue of employment is left out of the first condition (reg. 5(3)(a)(ii)), which reg. 4(2) lifts; reg. 5(5) counts a course at a school or college without that exclusion, so the apprentice has not "ceased to receive full-time education" and the reg. 5(4)(a) bar does not apply. The property oracle had the same assumption and is fixed with it. YAML Case 5 now expects the apprentice on a course to qualify (£3,455); new Case 17 keeps an apprentice who has left every course out. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Review round 2 of #2300: a programme-level terminal-date variable set for one person gives everyone else its default (true), not their date fact, so a student whose date fact was set false became eligible again. No default can keep the formula for people left unset, so the two programme variables now say so and tell users setting both to set the programme variable for everyone whose status matters. Invariant 5 is qualified, and a crossed test shows both the hazard and the documented use, for each programme. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
After main's benefit cap work, two cases asserted that a 16-year-old who has left education is not a Housing Benefit young person, so their DLA or armed forces independence payment does not lift the cap, and the exemption's docstring named this PR's issue (#2287) as the gap. Under reg. 4 they are a Child Benefit qualifying young person until the 31 August after their 16th birthday, and so a Housing Benefit young person (HB Regs 2006 reg. 19), whose benefit lifts the cap (reg. 75F(1)). Both expectations and the docstring now say so; Universal Credit's exemption is unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
No conflicts. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LwQq6HKeDniUizUaZcFw4Q
No conflicts. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LwQq6HKeDniUizUaZcFw4Q
is_on_income_support gives the benefit unit's calculated award only to whichever of the claimant and partner receives it, so the fragment says so. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LwQq6HKeDniUizUaZcFw4Q
The guard's docstring now says it looks for draws and not for the clock (#2875). The comments on the splitmix64 and Universal Credit entries say which constants the scan recognises, which entries reach the hash and how Simulation.__init__ applies the modifier. Two test rows follow PolicyEngine/policyengine-us#10447 at 221f2fe44, and a changelog fragment records the guard. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
inflation_adjustment and inflation_adjustment_ahc read the base year from gov.simulation.price_base_year instead of the clock, and inflation_adjustment_ahc uses one index on both dates. A code-health test fails on any clock read under policyengine_uk/variables. Fixes #2875 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01LwQq6KUHRad4GgGptdPAWP
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A function-local import overwrote a module-level one in the flat alias map, so `import math as datetime` inside a helper hid a module-level datetime.datetime.now(). Each name now resolves in the innermost scope that imports it, and a class body's imports reach only that class body. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This was referenced Nov 6, 2026
Contributor
Aggregate impact on certified data: passBase Checks: 0 broken by this change, 1 failing on both sides, 0 fixed. Moves flagged for review: 49 (thresholds in Notes
Moves flagged for review
9 more flags are in the report artifact. Receipts by tax head
Spending by benefit and service
Caseloads
Government balance, each component counted once
Model totals
Household income totals
Poverty rates
People in poverty
Inequality
Mean household net income in each side's own deciles
Standard reform costings (change in gov_balance)
Standard reforms: change in each reform's target
Household net income by decileMean change in household net income per household, and as a share of the decile's base income; households stay in the base run's deciles.
No change in any year: population. Checks failing on both base and head (1)
Counted-once checkThe check raised each of 92 components by £1 per unit in 2026-27 and read every head of its totals household by household: its own head must rise by exactly £1, no other tax head may move, no other benefit may rise, and the model's own total must move by what its heads moved. A response that is not finite breaks every rule. The check measures a head that breaks a rule twice more: once holding every other head at its baseline value, and once holding every other head at its value after the first £1 while the component rises again. The head fails where either measurement still shows the response. The check cannot hold a head at a value that is not finite. In each unit where another head is not finite after the first £1, it holds that head at its baseline value, and the second measurement judges no household the unit belongs to. There a head fails wherever it first broke a rule, or wherever the check could not confirm whether its response contains the component. 0 failed and the check skipped 0.
Provenance
Dataset: Receipts by tax head: head levels
Spending by benefit and service: head levels
Caseloads: head levels
Government balance, each component counted once: head levels
Model totals: head levels
Household income totals: head levels
Population: head levels
Poverty rates: head levels
People in poverty: head levels
Inequality: head levels
Mean household net income in each side's own deciles: head levels
Standard reform costings (change in gov_balance): head levels
Standard reforms: change in each reform's target: head levels
|
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I merged 14 approved pull requests onto main at
5c0850da5on this branch, so CI can test them as one tree. Please don't merge it: I'll close it and delete the branch when the run finishes, and post the result here first.I merged them in this order:
No pair of these has a text conflict. The order above, its reverse and four shuffled orders all end at the same tree,
fc40d74a4, so CI here tests what landing all 14 produces.I left out the other five approved pull requests. #1966, #2002, #2828 and #2845 each conflict with one or more of the 14, and #2010 contains #2007, which still needs a review.
Reproducer:
policyengine_uk/tests/test_price_base_year.py::test_reform_changes_the_base_year🤖 Generated with Claude Code