FAILURE MAP
← Case archive

FA-59696 / Subscription proration billing / Open access

Volume discount tiers on quantity changes: discounted unit rounding · case 01

Discounted prorations drift by cents proportional to the quantity.

Verified by executionVariant 1 · 8 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

The discounted unit price is rounded to cents before multiplying by quantity.

VERIFIED REPAIR

Restore the contract rule at the discounted unit rounding step: use `return q * x['unit'] * (100 - tier_pct(q))`.

Unsuccessful approach: The attempt truncates the discounted unit price instead, still rounding before quantity.

Case contract

Input {unit, old, new quantities, tiers: ascending [[min_qty, pct_off]], left, period}. A quantity's discount is the pct of the last tier with qty >= min_qty and applies to all units: total(q) = q*unit*(100-pct) in cent-percent. Proration of q = total(q)*left/(100*period) half-up. Return [-proration(old), proration(new)].

Why this case matters

Crossing a volume tier on a mid-cycle quantity change re-prices every unit for the rest of the period.

1 / The failure

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(x):
    def tier_pct(q):
        pct = 0
        for mn, off in x['tiers']:
            if q >= mn:
                pct = off
        return pct
    def total(q):
        return q * ((x['unit'] * (100 - tier_pct(q)) + 50) // 100) * 100
    def pr(q):
        return (total(q) * x['left'] * 2 + 100 * x['period']) // (200 * x['period'])
    return [-pr(x['old']), pr(x['new'])]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[('regression', {'unit': 1396, 'old': 18, 'new': 39, 'tiers': [[1, 0], [5, 10], [31, 25]], 'left': 21, 'period': 31}, [-15320, 27661]), ('regression', {'unit': 999, 'old': 18, 'new': 1, 'tiers': [[1, 0], [8, 10], [28, 15]], 'left': 25, 'period': 31}, [-13051, 806]), ('partial-repair probe', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('partial-repair probe', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('normal control', {'unit': 1500, 'old': 37, 'new': 19, 'tiers': [[1, 0], [5, 5], [28, 15]], 'left': 13, 'period': 31}, [-19783, 11354]), ('normal control', {'unit': 1500, 'old': 21, 'new': 45, 'tiers': [[1, 0], [7, 10], [29, 20]], 'left': 20, 'period': 31}, [-18290, 34839]), ('normal control', {'unit': 1000, 'old': 9, 'new': 22, 'tiers': [[1, 0], [8, 5], [22, 20]], 'left': 24, 'period': 31}, [-6619, 13626]), ('normal control', {'unit': 999, 'old': 1, 'new': 19, 'tiers': [[1, 0], [10, 5], [31, 25]], 'left': 1, 'period': 31}, [-32, 582])], [('regression', {'unit': 999, 'old': 18, 'new': 1, 'tiers': [[1, 0], [8, 10], [28, 15]], 'left': 25, 'period': 31}, [-13051, 806]), ('regression', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('partial-repair probe', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('partial-repair probe', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('normal control', {'unit': 1500, 'old': 38, 'new': 12, 'tiers': [[1, 0], [8, 10], [31, 25]], 'left': 15, 'period': 31}, [-20685, 7839]), ('normal control', {'unit': 1500, 'old': 22, 'new': 32, 'tiers': [[1, 0], [5, 5], [35, 25]], 'left': 9, 'period': 31}, [-9102, 13239]), ('normal control', {'unit': 1500, 'old': 6, 'new': 44, 'tiers': [[1, 0], [10, 10], [22, 15]], 'left': 22, 'period': 31}, [-6387, 39813]), ('normal control', {'unit': 1000, 'old': 48, 'new': 14, 'tiers': [[1, 0], [9, 5], [31, 20]], 'left': 21, 'period': 31}, [-26013, 9010])], [('regression', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('regression', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('partial-repair probe', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('partial-repair probe', {'unit': 999, 'old': 39, 'new': 26, 'tiers': [[1, 0], [7, 10], [32, 15]], 'left': 27, 'period': 31}, [-28844, 20360]), ('normal control', {'unit': 1500, 'old': 26, 'new': 40, 'tiers': [[1, 0], [7, 10], [26, 20]], 'left': 20, 'period': 31}, [-20129, 30968]), ('normal control', {'unit': 1000, 'old': 5, 'new': 17, 'tiers': [[1, 0], [5, 5], [37, 25]], 'left': 10, 'period': 31}, [-1532, 5210]), ('normal control', {'unit': 1000, 'old': 27, 'new': 46, 'tiers': [[1, 0], [9, 5], [20, 15]], 'left': 4, 'period': 31}, [-2961, 5045]), ('normal control', {'unit': 1000, 'old': 18, 'new': 17, 'tiers': [[1, 0], [10, 10], [32, 25]], 'left': 24, 'period': 31}, [-12542, 11845])], [('regression', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('regression', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('partial-repair probe', {'unit': 999, 'old': 9, 'new': 38, 'tiers': [[1, 0], [6, 5], [32, 15]], 'left': 27, 'period': 31}, [-7439, 28104]), ('partial-repair probe', {'unit': 999, 'old': 10, 'new': 1, 'tiers': [[1, 0], [5, 10], [22, 15]], 'left': 29, 'period': 31}, [-8411, 935]), ('normal control', {'unit': 1000, 'old': 41, 'new': 30, 'tiers': [[1, 0], [5, 5], [20, 15]], 'left': 23, 'period': 31}, [-25856, 18919]), ('normal control', {'unit': 1000, 'old': 1, 'new': 2, 'tiers': [[1, 0], [10, 5], [21, 20]], 'left': 30, 'period': 31}, [-968, 1935]), ('normal control', {'unit': 1500, 'old': 42, 'new': 42, 'tiers': [[1, 0], [9, 10], [33, 25]], 'left': 30, 'period': 31}, [-45726, 45726]), ('normal control', {'unit': 1500, 'old': 45, 'new': 35, 'tiers': [[1, 0], [9, 10], [25, 25]], 'left': 26, 'period': 31}, [-42460, 33024])], [('regression', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('regression', {'unit': 999, 'old': 39, 'new': 26, 'tiers': [[1, 0], [7, 10], [32, 15]], 'left': 27, 'period': 31}, [-28844, 20360]), ('partial-repair probe', {'unit': 999, 'old': 24, 'new': 6, 'tiers': [[1, 0], [5, 10], [35, 15]], 'left': 27, 'period': 31}, [-18794, 4699]), ('partial-repair probe', {'unit': 999, 'old': 37, 'new': 45, 'tiers': [[1, 0], [9, 10], [25, 20]], 'left': 31, 'period': 31}, [-29570, 35964]), ('normal control', {'unit': 999, 'old': 21, 'new': 1, 'tiers': [[1, 0], [10, 5], [31, 20]], 'left': 9, 'period': 31}, [-5786, 290]), ('normal control', {'unit': 1000, 'old': 27, 'new': 21, 'tiers': [[1, 0], [8, 10], [36, 15]], 'left': 24, 'period': 31}, [-18813, 14632]), ('normal control', {'unit': 1000, 'old': 4, 'new': 9, 'tiers': [[1, 0], [6, 10], [23, 20]], 'left': 30, 'period': 31}, [-3871, 7839]), ('normal control', {'unit': 1000, 'old': 26, 'new': 40, 'tiers': [[1, 0], [5, 5], [36, 25]], 'left': 25, 'period': 31}, [-19919, 24194])]]
for i, (label, args, expected) in enumerate(fixtures[N-1]):
    check("%s %d" % (label, i), solve(args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression 0[-15315, 27661][-15320, 27661]Failed
regression 1[-13050, 806][-13051, 806]Failed
partial-repair probe 2[-38652, 28421][-38646, 28424]Failed
partial-repair probe 3[-38212, 16503][-38214, 16502]Failed
normal control 4[-19783, 11354][-19783, 11354]Passed
normal control 5[-18290, 34839][-18290, 34839]Passed
normal control 6[-6619, 13626][-6619, 13626]Passed
normal control 7[-32, 582][-32, 582]Passed

SHA-256 / 47d81172c6f9d909707f987203f0778e1b4a0585bd262cf9dd61fe82c4da2f03

2 / The unsuccessful fix

Exit 1
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(x):
    def tier_pct(q):
        pct = 0
        for mn, off in x['tiers']:
            if q >= mn:
                pct = off
        return pct
    def total(q):
        return q * (x['unit'] * (100 - tier_pct(q)) // 100) * 100
    def pr(q):
        return (total(q) * x['left'] * 2 + 100 * x['period']) // (200 * x['period'])
    return [-pr(x['old']), pr(x['new'])]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[('regression', {'unit': 1396, 'old': 18, 'new': 39, 'tiers': [[1, 0], [5, 10], [31, 25]], 'left': 21, 'period': 31}, [-15320, 27661]), ('regression', {'unit': 999, 'old': 18, 'new': 1, 'tiers': [[1, 0], [8, 10], [28, 15]], 'left': 25, 'period': 31}, [-13051, 806]), ('partial-repair probe', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('partial-repair probe', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('normal control', {'unit': 1500, 'old': 37, 'new': 19, 'tiers': [[1, 0], [5, 5], [28, 15]], 'left': 13, 'period': 31}, [-19783, 11354]), ('normal control', {'unit': 1500, 'old': 21, 'new': 45, 'tiers': [[1, 0], [7, 10], [29, 20]], 'left': 20, 'period': 31}, [-18290, 34839]), ('normal control', {'unit': 1000, 'old': 9, 'new': 22, 'tiers': [[1, 0], [8, 5], [22, 20]], 'left': 24, 'period': 31}, [-6619, 13626]), ('normal control', {'unit': 999, 'old': 1, 'new': 19, 'tiers': [[1, 0], [10, 5], [31, 25]], 'left': 1, 'period': 31}, [-32, 582])], [('regression', {'unit': 999, 'old': 18, 'new': 1, 'tiers': [[1, 0], [8, 10], [28, 15]], 'left': 25, 'period': 31}, [-13051, 806]), ('regression', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('partial-repair probe', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('partial-repair probe', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('normal control', {'unit': 1500, 'old': 38, 'new': 12, 'tiers': [[1, 0], [8, 10], [31, 25]], 'left': 15, 'period': 31}, [-20685, 7839]), ('normal control', {'unit': 1500, 'old': 22, 'new': 32, 'tiers': [[1, 0], [5, 5], [35, 25]], 'left': 9, 'period': 31}, [-9102, 13239]), ('normal control', {'unit': 1500, 'old': 6, 'new': 44, 'tiers': [[1, 0], [10, 10], [22, 15]], 'left': 22, 'period': 31}, [-6387, 39813]), ('normal control', {'unit': 1000, 'old': 48, 'new': 14, 'tiers': [[1, 0], [9, 5], [31, 20]], 'left': 21, 'period': 31}, [-26013, 9010])], [('regression', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('regression', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('partial-repair probe', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('partial-repair probe', {'unit': 999, 'old': 39, 'new': 26, 'tiers': [[1, 0], [7, 10], [32, 15]], 'left': 27, 'period': 31}, [-28844, 20360]), ('normal control', {'unit': 1500, 'old': 26, 'new': 40, 'tiers': [[1, 0], [7, 10], [26, 20]], 'left': 20, 'period': 31}, [-20129, 30968]), ('normal control', {'unit': 1000, 'old': 5, 'new': 17, 'tiers': [[1, 0], [5, 5], [37, 25]], 'left': 10, 'period': 31}, [-1532, 5210]), ('normal control', {'unit': 1000, 'old': 27, 'new': 46, 'tiers': [[1, 0], [9, 5], [20, 15]], 'left': 4, 'period': 31}, [-2961, 5045]), ('normal control', {'unit': 1000, 'old': 18, 'new': 17, 'tiers': [[1, 0], [10, 10], [32, 25]], 'left': 24, 'period': 31}, [-12542, 11845])], [('regression', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('regression', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('partial-repair probe', {'unit': 999, 'old': 9, 'new': 38, 'tiers': [[1, 0], [6, 5], [32, 15]], 'left': 27, 'period': 31}, [-7439, 28104]), ('partial-repair probe', {'unit': 999, 'old': 10, 'new': 1, 'tiers': [[1, 0], [5, 10], [22, 15]], 'left': 29, 'period': 31}, [-8411, 935]), ('normal control', {'unit': 1000, 'old': 41, 'new': 30, 'tiers': [[1, 0], [5, 5], [20, 15]], 'left': 23, 'period': 31}, [-25856, 18919]), ('normal control', {'unit': 1000, 'old': 1, 'new': 2, 'tiers': [[1, 0], [10, 5], [21, 20]], 'left': 30, 'period': 31}, [-968, 1935]), ('normal control', {'unit': 1500, 'old': 42, 'new': 42, 'tiers': [[1, 0], [9, 10], [33, 25]], 'left': 30, 'period': 31}, [-45726, 45726]), ('normal control', {'unit': 1500, 'old': 45, 'new': 35, 'tiers': [[1, 0], [9, 10], [25, 25]], 'left': 26, 'period': 31}, [-42460, 33024])], [('regression', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('regression', {'unit': 999, 'old': 39, 'new': 26, 'tiers': [[1, 0], [7, 10], [32, 15]], 'left': 27, 'period': 31}, [-28844, 20360]), ('partial-repair probe', {'unit': 999, 'old': 24, 'new': 6, 'tiers': [[1, 0], [5, 10], [35, 15]], 'left': 27, 'period': 31}, [-18794, 4699]), ('partial-repair probe', {'unit': 999, 'old': 37, 'new': 45, 'tiers': [[1, 0], [9, 10], [25, 20]], 'left': 31, 'period': 31}, [-29570, 35964]), ('normal control', {'unit': 999, 'old': 21, 'new': 1, 'tiers': [[1, 0], [10, 5], [31, 20]], 'left': 9, 'period': 31}, [-5786, 290]), ('normal control', {'unit': 1000, 'old': 27, 'new': 21, 'tiers': [[1, 0], [8, 10], [36, 15]], 'left': 24, 'period': 31}, [-18813, 14632]), ('normal control', {'unit': 1000, 'old': 4, 'new': 9, 'tiers': [[1, 0], [6, 10], [23, 20]], 'left': 30, 'period': 31}, [-3871, 7839]), ('normal control', {'unit': 1000, 'old': 26, 'new': 40, 'tiers': [[1, 0], [5, 5], [36, 25]], 'left': 25, 'period': 31}, [-19919, 24194])]]
for i, (label, args, expected) in enumerate(fixtures[N-1]):
    check("%s %d" % (label, i), solve(args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression 0[-15315, 27661][-15320, 27661]Failed
regression 1[-13050, 806][-13051, 806]Failed
partial-repair probe 2[-38640, 28421][-38646, 28424]Failed
partial-repair probe 3[-38212, 16499][-38214, 16502]Failed
normal control 4[-19783, 11354][-19783, 11354]Passed
normal control 5[-18290, 34839][-18290, 34839]Passed
normal control 6[-6619, 13626][-6619, 13626]Passed
normal control 7[-32, 582][-32, 582]Passed

SHA-256 / b87831a39da779c0aafda10c1c42cc065909d8b37ea2b5624c6b1b665220c3c8

3 / The verified repair

Exit 0
"""Failure Map reference implementation. Python standard library only."""
import json

N = 1
observations = []
def solve(x):
    def tier_pct(q):
        pct = 0
        for mn, off in x['tiers']:
            if q >= mn:
                pct = off
        return pct
    def total(q):
        return q * x['unit'] * (100 - tier_pct(q))
    def pr(q):
        return (total(q) * x['left'] * 2 + 100 * x['period']) // (200 * x['period'])
    return [-pr(x['old']), pr(x['new'])]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
fixtures = [[('regression', {'unit': 1396, 'old': 18, 'new': 39, 'tiers': [[1, 0], [5, 10], [31, 25]], 'left': 21, 'period': 31}, [-15320, 27661]), ('regression', {'unit': 999, 'old': 18, 'new': 1, 'tiers': [[1, 0], [8, 10], [28, 15]], 'left': 25, 'period': 31}, [-13051, 806]), ('partial-repair probe', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('partial-repair probe', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('normal control', {'unit': 1500, 'old': 37, 'new': 19, 'tiers': [[1, 0], [5, 5], [28, 15]], 'left': 13, 'period': 31}, [-19783, 11354]), ('normal control', {'unit': 1500, 'old': 21, 'new': 45, 'tiers': [[1, 0], [7, 10], [29, 20]], 'left': 20, 'period': 31}, [-18290, 34839]), ('normal control', {'unit': 1000, 'old': 9, 'new': 22, 'tiers': [[1, 0], [8, 5], [22, 20]], 'left': 24, 'period': 31}, [-6619, 13626]), ('normal control', {'unit': 999, 'old': 1, 'new': 19, 'tiers': [[1, 0], [10, 5], [31, 25]], 'left': 1, 'period': 31}, [-32, 582])], [('regression', {'unit': 999, 'old': 18, 'new': 1, 'tiers': [[1, 0], [8, 10], [28, 15]], 'left': 25, 'period': 31}, [-13051, 806]), ('regression', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('partial-repair probe', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('partial-repair probe', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('normal control', {'unit': 1500, 'old': 38, 'new': 12, 'tiers': [[1, 0], [8, 10], [31, 25]], 'left': 15, 'period': 31}, [-20685, 7839]), ('normal control', {'unit': 1500, 'old': 22, 'new': 32, 'tiers': [[1, 0], [5, 5], [35, 25]], 'left': 9, 'period': 31}, [-9102, 13239]), ('normal control', {'unit': 1500, 'old': 6, 'new': 44, 'tiers': [[1, 0], [10, 10], [22, 15]], 'left': 22, 'period': 31}, [-6387, 39813]), ('normal control', {'unit': 1000, 'old': 48, 'new': 14, 'tiers': [[1, 0], [9, 5], [31, 20]], 'left': 21, 'period': 31}, [-26013, 9010])], [('regression', {'unit': 4294, 'old': 31, 'new': 18, 'tiers': [[1, 0], [10, 5], [21, 25]], 'left': 12, 'period': 31}, [-38646, 28424]), ('regression', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('partial-repair probe', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('partial-repair probe', {'unit': 999, 'old': 39, 'new': 26, 'tiers': [[1, 0], [7, 10], [32, 15]], 'left': 27, 'period': 31}, [-28844, 20360]), ('normal control', {'unit': 1500, 'old': 26, 'new': 40, 'tiers': [[1, 0], [7, 10], [26, 20]], 'left': 20, 'period': 31}, [-20129, 30968]), ('normal control', {'unit': 1000, 'old': 5, 'new': 17, 'tiers': [[1, 0], [5, 5], [37, 25]], 'left': 10, 'period': 31}, [-1532, 5210]), ('normal control', {'unit': 1000, 'old': 27, 'new': 46, 'tiers': [[1, 0], [9, 5], [20, 15]], 'left': 4, 'period': 31}, [-2961, 5045]), ('normal control', {'unit': 1000, 'old': 18, 'new': 17, 'tiers': [[1, 0], [10, 10], [32, 25]], 'left': 24, 'period': 31}, [-12542, 11845])], [('regression', {'unit': 4525, 'old': 44, 'new': 17, 'tiers': [[1, 0], [5, 5], [26, 15]], 'left': 7, 'period': 31}, [-38214, 16502]), ('regression', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('partial-repair probe', {'unit': 999, 'old': 9, 'new': 38, 'tiers': [[1, 0], [6, 5], [32, 15]], 'left': 27, 'period': 31}, [-7439, 28104]), ('partial-repair probe', {'unit': 999, 'old': 10, 'new': 1, 'tiers': [[1, 0], [5, 10], [22, 15]], 'left': 29, 'period': 31}, [-8411, 935]), ('normal control', {'unit': 1000, 'old': 41, 'new': 30, 'tiers': [[1, 0], [5, 5], [20, 15]], 'left': 23, 'period': 31}, [-25856, 18919]), ('normal control', {'unit': 1000, 'old': 1, 'new': 2, 'tiers': [[1, 0], [10, 5], [21, 20]], 'left': 30, 'period': 31}, [-968, 1935]), ('normal control', {'unit': 1500, 'old': 42, 'new': 42, 'tiers': [[1, 0], [9, 10], [33, 25]], 'left': 30, 'period': 31}, [-45726, 45726]), ('normal control', {'unit': 1500, 'old': 45, 'new': 35, 'tiers': [[1, 0], [9, 10], [25, 25]], 'left': 26, 'period': 31}, [-42460, 33024])], [('regression', {'unit': 999, 'old': 14, 'new': 7, 'tiers': [[1, 0], [8, 10], [27, 25]], 'left': 24, 'period': 31}, [-9745, 5414]), ('regression', {'unit': 999, 'old': 39, 'new': 26, 'tiers': [[1, 0], [7, 10], [32, 15]], 'left': 27, 'period': 31}, [-28844, 20360]), ('partial-repair probe', {'unit': 999, 'old': 24, 'new': 6, 'tiers': [[1, 0], [5, 10], [35, 15]], 'left': 27, 'period': 31}, [-18794, 4699]), ('partial-repair probe', {'unit': 999, 'old': 37, 'new': 45, 'tiers': [[1, 0], [9, 10], [25, 20]], 'left': 31, 'period': 31}, [-29570, 35964]), ('normal control', {'unit': 999, 'old': 21, 'new': 1, 'tiers': [[1, 0], [10, 5], [31, 20]], 'left': 9, 'period': 31}, [-5786, 290]), ('normal control', {'unit': 1000, 'old': 27, 'new': 21, 'tiers': [[1, 0], [8, 10], [36, 15]], 'left': 24, 'period': 31}, [-18813, 14632]), ('normal control', {'unit': 1000, 'old': 4, 'new': 9, 'tiers': [[1, 0], [6, 10], [23, 20]], 'left': 30, 'period': 31}, [-3871, 7839]), ('normal control', {'unit': 1000, 'old': 26, 'new': 40, 'tiers': [[1, 0], [5, 5], [36, 25]], 'left': 25, 'period': 31}, [-19919, 24194])]]
for i, (label, args, expected) in enumerate(fixtures[N-1]):
    check("%s %d" % (label, i), solve(args), expected)
print(json.dumps({"observations": observations, "passed": all(x["passed"] for x in observations)}, ensure_ascii=False))
raise SystemExit(0 if all(x["passed"] for x in observations) else 1)
Boundary fixtureActualExpectedOutcome
regression 0[-15320, 27661][-15320, 27661]Passed
regression 1[-13051, 806][-13051, 806]Passed
partial-repair probe 2[-38646, 28424][-38646, 28424]Passed
partial-repair probe 3[-38214, 16502][-38214, 16502]Passed
normal control 4[-19783, 11354][-19783, 11354]Passed
normal control 5[-18290, 34839][-18290, 34839]Passed
normal control 6[-6619, 13626][-6619, 13626]Passed
normal control 7[-32, 582][-32, 582]Passed

SHA-256 / dbf508407b7211694bc641f4a41f0c4cccf8cca9f24669abf97480e565d9ec11

Verification & scope

A deterministic teaching model of a stipulated billing rule. It makes no claim to reproduce any billing provider's exact behaviour and is not billing software. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.

Observations recorded using Python 3.12.14 at 2026-09-29T14:46:38.789694+00:00.

Case digest / 49e0c69aec64850495074b8cafb92d42e655485703f2876b15bd4ca2ad799f21