Reject Self-Referential Parameter

Review ask

As a clinical trial statistician or programmer, challenge anything in this benchmark: describe a problem in Comments, or propose a fix with the Edit button on any source file.

Task: Is this a specification error a study team could realistically make, and should the build stop on it rather than continue?
Inputs: Do they trigger the failure for the reason the summary gives?
Expected failure: Is this the right error, raised at the right point, and would it tell the programmer what to fix?

The other tabs beside Comments are background for assessment, not under review.

Summary

EditUnder reviewWhat this benchmark means

README.md

Goal: derive ADT, the analysis date, and AVAL, the analysis value, for each subject, collection date, and parameter: the alanine aminotransferase (ALT) and aspartate aminotransferase (AST) results keep their measured values, while a third ratio parameter (ASTALT, "Aspartate to Alanine Aminotransferase Ratio") divides the day's aspartate result by its alanine result.

Input: each input record is one collected result from LB, carrying its study, subject, and sequence number alongside LBDTC, the date the sample was collected; LBTESTCD, the test code; LBTEST, the test name; and LBSTRESN, the numeric result, which is missing when no numeric value was returned. One subject's only record is an aspartate result, and one alanine result is zero.

Variables:

ADT: the collection date, taken directly from LBDTC.
AVAL: the analysis value. On a transaminase record it repeats LBSTRESN. On the ratio record it divides the day's aspartate result by the alanine result. A ratio record exists for each alanine record, and its value is missing when that day has no aspartate record or either result has no value; a zero alanine result likewise leaves the ratio missing.

The ratio reads its two values from other records of the result being built, but those values are themselves the AVAL under definition, so no value can be assembled. The run is rejected before any data is read, and no artifact is accepted. The expected file shows what the completed records would contain, but no row is produced.

How to fix

First decide which run owns the ratio. A parameter whose value comes from other parameters belongs to a run that reads them as data, so build the two transaminase parameters first and take the ratio in a second specification that declares the completed dataset as one of its sources:

input:
  ADLB_RAW: adlb.csv

The transaminase results are then ordinary records with keys, and the ratio is an ordinary lookup of the aspartate result on the alanine record's date.

Do not reach the other parameter by counting rows instead. Which row sits one position away depends on how the records happen to be ordered, so a sample that answers correctly today stops doing so as soon as a third parameter is added.

Expected failure

Editthe run is rejected; no artifact is accepted

expected/error.yaml

phase: validation
condition: dependency_cycle
requirement: REQ-0072

expected/error.yaml

phase: validation
condition: dependency_cycle
spec_paths:
  - columns.ALTVAL.derivation.aggregate
  - columns.AVAL.derivation.case
requirement: REQ-0072
context:
  cycle: [ALTVAL, AVAL, ALTVAL]

YAML specification

1 spec file

spec.yaml

Edit104 lines

1schema_version: "1.0"2domain: ADLB3keys: [STUDYID, USUBJID, ADT, PARAMCD]4input:5  LB: {path: input/lb.csv, types: {LBSTRESN: float, LBDTC: date}}67output:8  path: adlb.csv9  columns: [STUDYID, USUBJID, ADT, PARAMCD, PARAM, AVAL]1011columns:12  - name: STUDYID13    type: str14    label: Study Identifier15    derivation: LB.STUDYID1617  - name: USUBJID18    type: str19    label: Unique Subject Identifier20    derivation: LB.USUBJID2122  - name: ADT23    type: date24    label: Analysis Date25    derivation: LB.LBDTC2627  - name: PARAMCD28    type: str29    label: Parameter Code30    verifications:31      allowed_values:32        values: [ALT, AST, ASTALT]3334  - name: PARAM35    type: str36    label: Parameter3738  - name: LBRESN39    type: float40    label: Numeric Laboratory Result4142  # The day's ALT and AST results, read back from the rows being built:43  # the largest AVAL among the subject's rows of that parameter on that date.44  # AVAL below is computed from them, which closes the cycle this benchmark45  # shows.46  - name: ALTVAL47    type: float48    label: Alanine Aminotransferase Analysis Value49    derivation:50      aggregate:51        filter: "PARAMCD = 'ALT'"52        group_by: [STUDYID, USUBJID, ADT]53        expr: "MAX(AVAL)"5455  - name: ASTVAL56    type: float57    label: Aspartate Aminotransferase Analysis Value58    derivation:59      aggregate:60        filter: "PARAMCD = 'AST'"61        group_by: [STUDYID, USUBJID, ADT]62        expr: "MAX(AVAL)"6364  # NULLIF leaves the ratio empty instead of dividing by a zero ALT.65  - name: AVAL66    type: float67    label: Analysis Value68    derivation:69      case:70        - when: "PARAMCD = 'ASTALT'"71          then:72            compute:73              expr: "ASTVAL / NULLIF(ALTVAL, 0)"74        - otherwise: LBRESN7576rows:77  - id: alt78    filter: "LB.LBTESTCD = 'ALT'"79    derivations:80      PARAMCD: {literal: ALT}81      PARAM: {literal: Alanine Aminotransferase}82      LBRESN: LB.LBSTRESN8384  - id: ast85    filter: "LB.LBTESTCD = 'AST'"86    derivations:87      PARAMCD: {literal: AST}88      PARAM: {literal: Aspartate Aminotransferase}89      LBRESN: LB.LBSTRESN9091  # One ratio row per ALT record; its value comes from AVAL above.92  - id: ast_alt_ratio93    filter: "LB.LBTESTCD = 'ALT'"94    derivations:95      PARAMCD: {literal: ASTALT}96      PARAM: {literal: Aspartate to Alanine Aminotransferase Ratio}97      LBRESN: {literal: null}9899verifications:100  - unique:101      columns: [STUDYID, USUBJID, ADT, PARAMCD]102  - row_count:103      min: 15104      max: 15

Inputs

Under review1 source file

input/lb.csv

Edit10 rows

input/lb.csv
STUDYID	LBSEQ	LBDTC	LBTESTCD	LBTEST	LBSTRESN
YAMAA-01	1	2025-03-03	ALT	Alanine Aminotransferase	40
YAMAA-01	2	2025-03-03	AST	Aspartate Aminotransferase	60
YAMAA-01	3	2025-03-17	ALT	Alanine Aminotransferase	25
YAMAA-01	4	2025-03-17	AST	Aspartate Aminotransferase	20
YAMAA-01	1	2025-03-04	ALT	Alanine Aminotransferase	30
YAMAA-01	1	2025-03-05	ALT	Alanine Aminotransferase	50
YAMAA-01	2	2025-03-05	AST	Aspartate Aminotransferase	—
YAMAA-01	1	2025-03-06	AST	Aspartate Aminotransferase	45
YAMAA-01	1	2025-03-07	ALT	Alanine Aminotransferase	0
YAMAA-01	2	2025-03-07	AST	Aspartate Aminotransferase	40

Unexpected Output

Under review15 rows presented to the failing check - not an accepted artifact

expected/adlb.csv

Edit15 rows

expected/adlb.csv
STUDYID	ADT	PARAMCD	PARAM	AVAL
YAMAA-01	2025-03-03	ALT	Alanine Aminotransferase	40
YAMAA-01	2025-03-17	ALT	Alanine Aminotransferase	25
YAMAA-01	2025-03-04	ALT	Alanine Aminotransferase	30
YAMAA-01	2025-03-05	ALT	Alanine Aminotransferase	50
YAMAA-01	2025-03-07	ALT	Alanine Aminotransferase	0
YAMAA-01	2025-03-03	AST	Aspartate Aminotransferase	60
YAMAA-01	2025-03-17	AST	Aspartate Aminotransferase	20
YAMAA-01	2025-03-05	AST	Aspartate Aminotransferase	—
YAMAA-01	2025-03-06	AST	Aspartate Aminotransferase	45
YAMAA-01	2025-03-07	AST	Aspartate Aminotransferase	40
YAMAA-01	2025-03-03	ASTALT	Aspartate to Alanine Aminotransferase Ratio	1.5
YAMAA-01	2025-03-17	ASTALT	Aspartate to Alanine Aminotransferase Ratio	0.8
YAMAA-01	2025-03-04	ASTALT	Aspartate to Alanine Aminotransferase Ratio	—
YAMAA-01	2025-03-05	ASTALT	Aspartate to Alanine Aminotransferase Ratio	—
YAMAA-01	2025-03-07	ASTALT	Aspartate to Alanine Aminotransferase Ratio	—

Shaded columns are derived— = missing value

Comments and assessment materials

Dataset Name	Variable Name	Variable Label	Type	Codelist/Controlled Terms	Computational Method	Origin
ADLB	STUDYID	Study Identifier	character		Copy value from LB.STUDYID.	Predecessor
ADLB	USUBJID	Unique Subject Identifier	character		Copy value from LB.USUBJID.	Predecessor
ADLB	ADT	Analysis Date	numeric	ISO 8601	Copy value from LB.LBDTC.	Predecessor
ADLB	PARAMCD	Parameter Code	character	ALT, AST, ASTALT	Set per constructed row; see the Row construction sheet.	Derived
ADLB	PARAM	Parameter	character		Set per constructed row; see the Row construction sheet.	Derived
ADLB	AVAL	Analysis Value	numeric		If PARAMCD = 'ASTALT' then compute ASTVAL / NULLIF(ALTVAL, 0); otherwise "LBRESN".	Derived

Template	Include when	PARAMCD	PARAM	LBRESN
alt	LB.LBTESTCD = 'ALT'	Constant "ALT"	Constant "Alanine Aminotransferase"	Copy LB.LBSTRESN
ast	LB.LBTESTCD = 'AST'	Constant "AST"	Constant "Aspartate Aminotransferase"	Copy LB.LBSTRESN
ast_alt_ratio	LB.LBTESTCD = 'ALT'	Constant "ASTALT"	Constant "Aspartate to Alanine Aminotransferase Ratio"	Constant "None"

Variable	Permitted values
PARAMCD	ALT, AST, ASTALT

Version	Date	Author	Description
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

How to fix

input/lb.csv

expected/adlb.csv

Comments

Mapping spec