Q37, run 2026-11-04-2022-sonnet

← back to the run

How many distinct legal entities does the data describe?

Tests: Identity: records with the same LEI are one entity.

Expected: count = 497

❌ R1 csv run 1 graded by code

511. There are 540 entity rows, and 29 are RESEARCH duplicates sharing an LEI with another row.

JSON: {"count": 511}

Why: count: expected 497, got 511

✅ R1 csv run 2 graded by code

497. That is 540 rows, less 29 duplicate LEIs, less 14 QA records (E16, E1340–E1351, E5170). Excluding QA is my assumption.

JSON: {"count": 497}

❌ R1 csv run 3 graded by code

511. There are 540 rows, minus 29 RESEARCH duplicates sharing an LEI with another row.

JSON: {"count": 511}

Why: count: expected 497, got 511

✅ R1 csv run 4 graded by code

497. The file has 540 rows, of which 29 are duplicate records (same LEI) and 14 are QA-source records that I excluded.

JSON: {"count": 497}

✅ R1 csv run 5 graded by code

497. There are 540 rows, minus 29 RESEARCH duplicates sharing an LEI, minus 14 QA rows (E16, E1340–E1351, E5170). Including QA rows gives 511.

JSON: {"count": 497}

✅ R4 csv+markdown run 1 graded by code

497. This is 16 core entities, 315 in the E1xxx range and 166 in the E5xxx range, after merging same-LEI duplicates and dropping QA records.

JSON: {"count": 497}

✅ R4 csv+markdown run 2 graded by code

497. There are 540 records; I removed 14 QA records and 29 duplicate records that share another record's LEI.

JSON: {"count": 497}

✅ R4 csv+markdown run 3 graded by code

497. This is 540 records, less 14 QA records and 29 duplicate LEI records. The two sovereigns have no LEI and are counted as entities.

JSON: {"count": 497}

✅ R4 csv+markdown run 4 graded by code

497. That is 540 records, minus 29 same-LEI duplicates and 14 QA records, and it includes the two sovereigns.

JSON: {"count": 497}

✅ R4 csv+markdown run 5 graded by code

497. That is 540 records, less 14 QA records and 29 duplicate-LEI records.

JSON: {"count": 497}

✅ R10 executed run 1 graded by code

497. Entities.csv has 497 rows after the merges, counting 16 + 315 + 166 across the E0x, E1xxx and E5xxx series.

JSON: {"count": 497}

✅ R10 executed run 2 graded by code

497 distinct entities in entities.csv, with duplicates already merged.

JSON: {"count": 497}

✅ R10 executed run 3 graded by code

497. This is 16 core rows, 315 E1xxx rows and 166 E5xxx rows, taking duplicates as already merged. I derived the E1xxx count by subtracting the 24 merged-away IDs from E1001–E1339.

JSON: {"count": 497}

✅ R10 executed run 4 graded by code

497 entity rows in entities.csv, after merging and including the two sovereigns.

JSON: {"count": 497}

✅ R10 executed run 5 graded by code

497 rows in entities.csv, where duplicates are already merged.

JSON: {"count": 497}

✅ R8 csv+axioms run 1 graded by code

497. This is 540 records, minus 14 QA records and 29 research duplicates that share a LEI. The two sovereigns are included.

JSON: {"count": 497}

✅ R8 csv+axioms run 2 graded by code

497. That is 540 records, less 14 QA records and 29 duplicate-LEI records.

JSON: {"count": 497}

✅ R8 csv+axioms run 3 graded by code

497. This is 540 rows, minus 14 QA records and 29 research duplicates sharing a LEI.

JSON: {"count": 497}

✅ R8 csv+axioms run 4 graded by code

497. That is 540 records, less 14 QA test records and 29 research duplicates matched by LEI.

JSON: {"count": 497}

✅ R8 csv+axioms run 5 graded by code

497. That is 540 records, less 14 QA records and 29 research duplicates sharing an LEI.

JSON: {"count": 497}