Skip to content

Commit 68a9954

Browse files
Nick  VaccarelloNick  Vaccarello
authored andcommitted
docs(NEXT_STEPS): mark dataset scaffolding complete; note CI validator and EIG tests as done
1 parent 962e7c2 commit 68a9954

File tree

1 file changed

+8
-8
lines changed

1 file changed

+8
-8
lines changed

medical_diagnosis_model/NEXT_STEPS.md

Lines changed: 8 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -68,11 +68,11 @@ Acceptance:
6868
- Version tag and changelog updated when schema changes.
6969
- Dataset scaffolding (v0)
7070
- Acceptance criteria:
71-
- JSON Schema for clinical cases present at `data/case.schema.json`.
72-
- Canonical dictionaries at `data/dictionaries/{symptoms.json,diseases.json}`.
73-
- Sample JSONL at `data/samples/cases_v0.1.jsonl` following the schema.
74-
- Validator CLI `data/validate_cases.py` passes on samples; `jsonschema` added to requirements.
75-
- Command: `python data/validate_cases.py data/samples/cases_v0.1.jsonl` prints `Validation passed: 0 errors`.
71+
- [x] JSON Schema for clinical cases present at `data/case.schema.json`.
72+
- [x] Canonical dictionaries at `data/dictionaries/{symptoms.json,diseases.json}`.
73+
- [x] Sample JSONL at `data/samples/cases_v0.1.jsonl` following the schema.
74+
- [x] Validator CLI `data/validate_cases.py` passes on samples; `jsonschema` added to requirements.
75+
- [x] Command: `python data/validate_cases.py data/samples/cases_v0.1.jsonl` prints `Validation passed: 0 errors`.
7676
- References: `data/README.md`, `data/case.schema.json`, `data/dictionaries/`, `data/samples/`, `data/validate_cases.py`.
7777
- Label policy
7878
- For each disease: criteria for “confirmed” (ICD‑10 + test) vs “presumptive”; clinician adjudication rules.
@@ -162,9 +162,9 @@ Acceptance:
162162
- Data dictionary (fields, ranges, encodings); update legal/privacy notes.
163163
- Unit tests for rules and pipeline; lightweight CI (lint + tests).
164164
- Acceptance criteria:
165-
- `docs/data_dictionary.md` added; CI workflow runs lint + unit tests on PRs.
166-
- EIG selector unit tests validate entropy reduction and expected next‑question behavior on toy distributions.
167-
- CI executes `python medical_diagnosis_model/data/validate_cases.py` against `data/samples/*.jsonl` and any tracked dataset JSONL files.
165+
- [ ] `docs/data_dictionary.md` added; CI workflow runs lint + unit tests on PRs.
166+
- [x] EIG selector unit tests validate entropy reduction and expected next‑question behavior on toy distributions.
167+
- [x] CI executes `python medical_diagnosis_model/data/validate_cases.py` against `data/samples/*.jsonl` and any tracked dataset JSONL files.
168168

169169
<a id="rag"></a>
170170

0 commit comments

Comments
 (0)