Column Composition

Review ask

As a clinical trial statistician or programmer, answer two short questions in Your review: the task and the inputs and outputs. Agreeing with both approves this benchmark; disagreeing with either needs a comment. Propose a fix with the Edit button on any source file, or add it to your comments.

We suggest spending no more than 5 minutes on a review. The other tabs beside Your review are background for assessment, not under review.

Summary

EditUnder reviewWhat this benchmark means

README.md

Public entries: Python and R. Passing coverage requires an installed inventory.

Goal: demonstrate a study level changing details inside inherited columns - AVAL, ANRIND, and PARCAT1.

Input: three spec files, read root-first:

spec_organization.yaml turns the laboratory test code, result, and reference range indicator into PARAMCD, AVAL, ANRIND, and PARCAT1, reading the indicator without regard to letter case; it marks AVAL as the analysis value and checks that it is not negative;
spec_compound.yaml adds the low reference range wording the organization does not list, and notes that AVAL comes from the central laboratory;
spec_study.yaml is the entry: it applies one corrected result, marks AVAL as adjudicated and checks that it is never missing, leaves an unrecognized indicator empty, states PARCAT1 as the one category this study collects, and declares the complete output.

Each level names the one level above it, so the chain reads organization, then compound, then study, and a later level wins where two levels state the same detail.

Note: a later level composes a column detail by detail. The added reference range wording joins the two the organization already lists, the three annotations on AVAL accumulate, and the case-insensitive reading survives a level that never mentions it. A level naming a different way to produce a value replaces it whole instead, which is why PARCAT1 stops reading the test code. Lists never accumulate: the study's own check on AVAL replaces the organization's.

YAML specification

3 spec files

spec_organization.yaml
Edit61 lines
1yamaa: "1.0"2type: spec3input:4  LB:5    path: input/lb.csv6    types:7      USUBJID: str8      LBTESTCD: str9      LBSTRESN: float10      LBNRIND: str11base: LB1213columns:14  - name: USUBJID15    type: str16    label: Unique Subject Identifier17    derivation: LB.USUBJID1819  - name: PARAMCD20    type: str21    label: Parameter Code22    derivation: LB.LBTESTCD2324  # The compound and study levels each add a metadata entry; the study25  # level also replaces the derivation and the whole verification list.26  - name: AVAL27    type: float28    label: Analysis Value29    derivation: LB.LBSTRESN30    verifications:31      - range:32          min: 033    metadata:34      analysis_role: analysis-value3536  # Later levels add an entry to dict and the mapping handlers; source and37  # case_sensitive are inherited from here unchanged.38  - name: ANRIND39    type: str40    label: Analysis Reference Range Indicator41    derivation:42      mapping:43        source: LB.LBNRIND44        dict:45          N: NORMAL46          H: HIGH47        case_sensitive: false4849  # The study level replaces this mapping with a literal.50  - name: PARCAT151    type: str52    label: Parameter Category53    derivation:54      mapping:55        source: LB.LBTESTCD56        dict:57          ALT: HEPATIC58          AST: HEPATIC5960metadata:61  scope: organization
spec_compound.yaml
Edit20 lines
1yamaa: "1.0"2type: spec3parent: spec_organization.yaml45columns:6  # Merged into the organization's AVAL: this entry joins analysis_role.7  - name: AVAL8    metadata:9      source_system: central-laboratory1011  # Merged into the organization's mapping: L joins N and H, and the12  # source and case_sensitive setting carry over.13  - name: ANRIND14    derivation:15      mapping:16        dict:17          L: LOW1819metadata:20  scope: compound
spec_study.yaml
Edit40 lines
1yamaa: "1.0"2type: spec3model: ADaM4parent: spec_compound.yaml5domain: ADLB6keys: [USUBJID, PARAMCD]78output:9  path: adlb.csv10  columns: [USUBJID, PARAMCD, AVAL, ANRIND, PARCAT1]1112columns:13  # A case is a different operation from the inherited plain source, so it14  # replaces it whole. The verification list replaces the organization's15  # range check (lists never merge); the metadata entry joins the other two.16  - name: AVAL17    derivation:18      case:19        - when: "USUBJID = '01' AND PARAMCD = 'AST'"20          then: 31.021        - otherwise: LB.LBSTRESN22    verifications:23      - not_missing24    metadata:25      review_status: adjudicated2627  # Adds only the handlers to the inherited mapping: an absent indicator28  # and one with no dict entry (such as x) both come out missing.29  - name: ANRIND30    derivation:31      mapping:32        missing: null33        unmapped: null3435  # A literal replaces the inherited mapping whole.36  - name: PARCAT137    derivation: "'CHEMISTRY'"3839metadata:40  scope: study

Inputs

Under review1 source file

input/lb.csv

Edit4 rows

input/lb.csv
LBTESTCD	LBSTRESN	LBSTRESU	LBNRIND
ALT	12.5	U/L	n
AST	-1.0	U/L	h
ALT	21.0	U/L	l
AST	30.0	U/L	x

Expected output

Under review4 expected rows

expected/adlb.csv

Edit4 rows

expected/adlb.csv
PARAMCD	AVAL	ANRIND	PARCAT1
ALT	12.5	NORMAL	CHEMISTRY
AST	31	HIGH	CHEMISTRY
ALT	21	LOW	CHEMISTRY
AST	30	—	CHEMISTRY

Shaded columns are derived— = missing value

Human review and assessment

Dataset Name	Variable Name	Computational Method	Origin
ADLB	AVAL	If USUBJID = '01' AND PARAMCD = 'AST' then "31.0"; otherwise "LB.LBSTRESN".	Derived
ADLB	ANRIND	Recode None: ; missing or unlisted values stay missing.	Derived
ADLB	PARCAT1	Set constant value "CHEMISTRY".	Assigned

Version	Date	Author	Description
1.7	2026-10-20	yamaa docs build	Read a specification in the 1.0 surface as its 0.5 counterpart; state that missing values stay missing where a mapping declares no missing handler, and leave out a declared missing answer that restates a handler's default
1.6	2026-10-17	yamaa docs build	Describe collected date parts written as ISO 8601 date text and times of day counted in seconds since midnight
1.5	2026-10-17	yamaa docs build	Name collected precision in words instead of letter codes; state imputation fallbacks, minimum source precision, bounds ignored outside the collected period, and invalid text with no handler as an error; describe date and datetime shifts, imputation flag codes, and partial text giving no date part
1.4	2026-10-15	yamaa docs build	State datetime precision codes through the hour and minute, and complete only the time parts a collected datetime lacks
1.3	2026-09-29	yamaa docs build	Show study codelist names, external dictionaries, and ISO 8601 formats in variable rows; list the document's codelist values and metadata
1.2	2026-09-29	yamaa docs build	Classify ADaM copies from a declared input as Predecessor and recodes as Derived, and SDTM recodes of locally derived columns as Derived
1.1	2026-09-29	yamaa docs build	Resolve lookups into the donor record they select, and give every derivation shape a plain-language rule instead of its raw YAML
1.0		yamaa docs build	Initial generation from spec.yaml

R

Edit3 lines

1library(yamaa)23yamaa_spec("spec_study.yaml")$run()$output

Python

Edit3 lines

1import yamaa23yamaa.spec("spec_study.yaml").run().output

input/lb.csv

expected/adlb.csv

Your review

Mapping spec

yamaa code

R

Python