FA-99961 / Evaluation metric aggregation / Open access
Support weighted recall divides by label count · case 01
Support weighted recall divides by label count.
ROOT CAUSE
The weighted sum uses the number of classes instead of total support.
THE FAILURE
The weighted sum uses the number of classes instead of total support.
Unsuccessful approach: Weighting by predicted counts shifts emphasis toward overpredicted labels.
Case contract
Return support-weighted recall for supplied labels; labels without support contribute zero; no support returns 0.
Why this case matters
A small explicit evaluation contract makes aggregation behavior reproducible across labeled examples.
1 / The failure
Exit 1"""Failure Map reference implementation. Python standard library only."""
import json
N = 1
observations = []
def solve(truth, pred, labels):
return sum(sum(a==k and b==k for a,b in zip(truth,pred))/sum(a==k for a in truth)*sum(a==k for a in truth) for k in labels)/len(labels) if labels and truth else 0
def check(label, actual, expected):
observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('support weighted',solve([0,0,0,1],[0,1,1,1],[0,1]),.5)
check('perfect',solve([0,1],[0,1],[0,1]),1)
check('all wrong',solve([0,1],[1,0],[0,1]),0)
check('empty labels',solve([0],[0],[]),0)
check('empty data',solve([],[],[0,1]),0)
check('single class',solve([N]*N,[N]*N,[N]),1)
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 |
|---|---|---|---|
| support weighted | 1.0 | 0.5 | Failed |
| perfect | 1.0 | 1 | Passed |
| all wrong | 0.0 | 0 | Passed |
| empty labels | 0 | 0 | Passed |
| empty data | 0 | 0 | Passed |
| single class | 1.0 | 1 | Passed |
SHA-256 / 30dfa707be8aac2097c6e75ac715e5330cbbb3aab69107b2f6d82380eb42a339
2 / The unsuccessful fix
Exit 1"""Failure Map reference implementation. Python standard library only."""
import json
N = 1
observations = []
def solve(truth, pred, labels):
return sum((sum(a==k and b==k for a,b in zip(truth,pred))/sum(a==k for a in truth) if sum(a==k for a in truth) else 0)*sum(b==k for b in pred) for k in labels)/len(truth) if truth else 0
def check(label, actual, expected):
observations.append({"check": label, "actual": actual, "expected": expected, "passed": actual == expected})
check('support weighted',solve([0,0,0,1],[0,1,1,1],[0,1]),.5)
check('perfect',solve([0,1],[0,1],[0,1]),1)
check('all wrong',solve([0,1],[1,0],[0,1]),0)
check('empty labels',solve([0],[0],[]),0)
check('empty data',solve([],[],[0,1]),0)
check('single class',solve([N]*N,[N]*N,[N]),1)
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 |
|---|---|---|---|
| support weighted | 0.8333333333333334 | 0.5 | Failed |
| perfect | 1.0 | 1 | Passed |
| all wrong | 0.0 | 0 | Passed |
| empty labels | 0.0 | 0 | Passed |
| empty data | 0 | 0 | Passed |
| single class | 1.0 | 1 | Passed |
SHA-256 / 556cab67d59d138a74afce695e71cd3df8894c1d2e4b250aae24fc521a1cd22b
HELD IN THE MEMBER ARCHIVE
The verified repair and its recorded checks are member-only.
This mechanism has 6 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.
Sign in to the archive ↗Verification & scope
Inputs are finite toy labels and scores; this defines no benchmark protocol, production metric library, or model quality claim. 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 2026-09-30T16:48:30.355570+00:00.
Case digest / ffed7376318af49ff220b0adb1aa03f67990db8e1c2c41b4a45c0f2163d82ee3