FA-100361 / Feature preprocessing fit scope / Open access

Population standardization uses held-out rows while fitting · case 01

Population standardization uses held-out rows while fitting.

Verified by executionVariant 1 · 7 checks per implementationDownload source bundle ↓JSON ↗

ROOT CAUSE

Fitting reads values outside the designated training rows.

THE FAILURE

Fitting reads values outside the designated training rows.

Unsuccessful approach: Filtering rows after computing the statistic still leaves held-out influence.

Case contract

Fit population mean and population standard deviation on nonempty train; transform values as z scores rounded to six decimals. A zero fitted deviation maps values to zero.

Why this case matters

A small deterministic preprocessing contract makes the fit boundary and transform behavior directly inspectable.

1 / The failure

Exit 1

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

N = 1
observations = []
def solve(train, values):
    if not values: return []
    m=sum(values)/len(values); s=(sum((x-m)**2 for x in values)/len(values))**0.5; return [round((x-m)/s,6) if s else 0 for x in values]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
a=[N,N+2]; check('population scale',solve([0,2],[1,3]),[0,2])
check('heldout does not change center',solve([0,2],[2]),[1])
check('center at mean',solve(a,[N+1]),[0])
check('constant training',solve([N,N],[N+3]),[0])
check('singleton training',solve([N],[N]),[0])
check('empty transform',solve([0,2],[]),[])
check('asymmetric',solve([0,4],[2]),[0])
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 fixture	Actual	Expected	Outcome
population scale	`[-1.0, 1.0]`	`[0, 2]`	Failed
heldout does not change center	`[0]`	`[1]`	Failed
center at mean	`[0]`	`[0]`	Passed
constant training	`[0]`	`[0]`	Passed
singleton training	`[0]`	`[0]`	Passed
empty transform	`[]`	`[]`	Passed
asymmetric	`[0]`	`[0]`	Passed

SHA-256 / 53d634d7a83500b0226214e4d45d4f4f5a274c06913fc0b9a9b76ee0bf148fa0

2 / The unsuccessful fix

Exit 1

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

N = 1
observations = []
def solve(train, values):
    v=train+values; m=sum(v)/len(v); s=(sum((x-m)**2 for x in v)/len(v))**0.5; return [round((x-m)/s,6) if s else 0 for x in values]
def check(label, actual, expected):
    observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
a=[N,N+2]; check('population scale',solve([0,2],[1,3]),[0,2])
check('heldout does not change center',solve([0,2],[2]),[1])
check('center at mean',solve(a,[N+1]),[0])
check('constant training',solve([N,N],[N+3]),[0])
check('singleton training',solve([N],[N]),[0])
check('empty transform',solve([0,2],[]),[])
check('asymmetric',solve([0,4],[2]),[0])
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 fixture	Actual	Expected	Outcome
population scale	`[-0.447214, 1.341641]`	`[0, 2]`	Failed
heldout does not change center	`[0.707107]`	`[1]`	Failed
center at mean	`[0.0]`	`[0]`	Passed
constant training	`[1.414214]`	`[0]`	Failed
singleton training	`[0]`	`[0]`	Passed
empty transform	`[]`	`[]`	Passed
asymmetric	`[0.0]`	`[0]`	Passed

SHA-256 / 34202b595e34f6b57aa2159f9783debc0baa588d503a318ed4a0708021d4dea3

HELD IN THE MEMBER ARCHIVE

The verified repair and its recorded checks are member-only.

This mechanism has 7 recorded checks per implementation. The open-access tier publishes the failure and the unsuccessful fix; the repaired source that passes every check, and the observations that prove it, are available to members.

Every case sharing this mechanism uses the same contract and the same repair, so this one record is held back for all of them.

Member access is invitation-based. Sign in with your invited account to inspect the repair.

Verification & scope

Toy scalar standardizer; it uses population variance and defines constant-feature output as zero. 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 2026-09-30T16:48:34.907773+00:00.

Case digest / f5488ec9ffb4edbcb3a32daedf4bf98281cc0727da46d3f144548f681d358959