As a clinical trial statistician or programmer, answer three short questions in Your review: the task, the inputs and outputs, and the AI-assisted review. Agreeing with all three approves this benchmark; disagreeing with any 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
README.mdPublic entries: Python and R. Passing coverage requires an installed inventory.
Goal: keep the investigator comment in CMNT and whether one
was collected in CMNTFL.
Input: demographics (DM) rows carrying the COMMENT field.
Variables:
CMNTis the comment exactly as the investigator recorded it: a comma inside it, quotation marks around a subject's own words, and a line break in the middle of it all reach the result unchanged; it is absent when no comment was collected.CMNTFLisYwhen a comment was collected andNwhen none was.
Note: a comment field that is empty, quoted or not, or holds
only spaces counts as no comment collected: the flag is N and the
comment is absent.
Inputs
Under review1 source fileinput/dm.csv
Edit6 rows| STUDYID | USUBJID | COMMENT |
|---|---|---|
| YAMAA-01 | Dose reduced, per protocol | |
| YAMAA-01 | Subject said "felt fine" | |
| YAMAA-01 | Visit missed rescheduled | |
| YAMAA-01 | — | |
| YAMAA-01 | — | |
| YAMAA-01 |
Expected output
Under review6 expected rowsexpected/adsl.csv
Edit6 rows| STUDYID | USUBJID | CMNT | CMNTFL |
|---|---|---|---|
| YAMAA-01 | Dose reduced, per protocol | Y | |
| YAMAA-01 | Subject said "felt fine" | Y | |
| YAMAA-01 | Visit missed rescheduled | Y | |
| YAMAA-01 | — | N | |
| YAMAA-01 | — | N | |
| YAMAA-01 | — | N |
Human review and assessment
| Dataset Name | Variable Name | Variable Label | Type | Codelist/Controlled Terms | Core | Computational Method | Origin |
|---|---|---|---|---|---|---|---|
| ADSL | STUDYID | Study Identifier | character | Copy value from DM.STUDYID. | Predecessor | ||
| ADSL | USUBJID | Unique Subject Identifier | character | Copy value from DM.USUBJID. | Predecessor | ||
| ADSL | CMNT | Investigator Comment | character | If str_contains(DM.COMMENT, '\S') then "DM.COMMENT"; otherwise blank. | Derived | ||
| ADSL | CMNTFL | Comment Collected Flag | character | If CMNT IS NULL then "N"; otherwise "Y". | Derived |
| Version | Date | Author | Description | Reviewer | Sign-off |
|---|---|---|---|---|---|
| 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 |
prompts/adam-adsl-investigator-comment/full.md
Edit16 linesFollowing CDISC ADaM standards, use the provided DM dataset to create an ADSL dataset with one record per subject.
The output dataset should contain the following columns in this order: STUDYID, USUBJID, CMNT, CMNTFL
CMNT is the comment exactly as the investigator recorded it: a comma inside it, quotation marks around a subject's own words, and a line break in the middle of it all reach the result unchanged, and CMNT has no value when no comment was collected. CMNTFL is "Y" when a comment was collected and "N" when none was. A comment field that is empty, quoted or not, or holds only spaces counts as no comment collected: CMNTFL is "N" and CMNT has no value.
Read the source datasets from /app/input and save the completed dataset as /app/output/adsl.csv.
prompts/adam-adsl-investigator-comment/conventions.md
Edit13 linesFollowing CDISC ADaM standards, use the provided DM dataset to create an ADSL dataset with one record per subject.
The output dataset should contain the following columns in this order: STUDYID, USUBJID, CMNT, CMNTFL
CMNT is the comment exactly as the investigator recorded it: a comma inside it, quotation marks around a subject's own words, and a line break in the middle of it all reach the result unchanged, and CMNT has no value. CMNTFL is "Y" or "N".
Read the source datasets from /app/input and save the completed dataset as /app/output/adsl.csv.
prompts/adam-adsl-investigator-comment/brief.md
Edit8 linesFollowing CDISC ADaM standards, use the provided DM dataset to create an ADSL dataset with one record per subject.
The output dataset should contain the following columns in this order: STUDYID, USUBJID, CMNT, CMNTFL
Read the source datasets from /app/input and save the completed dataset as /app/output/adsl.csv.
R
Edit3 lineslibrary(yamaa)yamaa_spec("spec.yaml")$run()$outputPython
Edit3 linesimport yamaayamaa.spec("spec.yaml").run().outputsolutions/adam-adsl-investigator-comment/result.R
Edit27 lines# Reference solution for the yamaa benchmark adam-adsl-investigator-comment (R track).## Written from the benchmark's full prompt and its inputs alone, as an agent# would write it. Harbor's oracle agent runs it to check that the prompt can# be solved and that the grader scores a correct answer 1.library(dplyr, warn.conflicts = FALSE)library(readr)dm <- read_csv( "/app/input/dm.csv", col_types = cols(.default = col_character()), trim_ws = FALSE)# A comment field that is empty, quoted or not, or holds only spaces# counts as no comment collected; any other comment is kept as recorded.adsl <- dm |> mutate( CMNT = if_else(grepl("\\S", COMMENT), COMMENT, NA_character_), CMNTFL = if_else(is.na(CMNT), "N", "Y") ) |> select(STUDYID, USUBJID, CMNT, CMNTFL) |> arrange(USUBJID)dir.create("/app/output", showWarnings = FALSE)write_csv(adsl, "/app/output/adsl.csv", na = "")solutions/adam-adsl-investigator-comment/result.py
Edit56 lines# Reference solution for the yamaa benchmark adam-adsl-investigator-comment (Python track).## Written from the benchmark's full prompt and its inputs alone, as an agent# would write it. Harbor's oracle agent runs it to check that the prompt can# be solved and that the grader scores a correct answer 1."""Create ADSL (one record per subject) from DM.CMNT carries the investigator comment exactly as recorded; CMNTFL is"Y" when a comment was collected and "N" when none was. A commentfield that is empty (quoted or not) or holds only spaces counts as nocomment."""import csvimport osINPUT_DIR = "/app/input"OUTPUT_DIR = "/app/output"FIELDNAMES = ["STUDYID", "USUBJID", "CMNT", "CMNTFL"]def main(): with open( os.path.join(INPUT_DIR, "dm.csv"), newline="", encoding="utf-8" ) as handle: dm_rows = list(csv.DictReader(handle)) adsl_rows = [] for row in dm_rows: comment = row.get("COMMENT") or "" collected = comment.strip() != "" adsl_rows.append( { "STUDYID": row["STUDYID"], "USUBJID": row["USUBJID"], "CMNT": comment if collected else "", "CMNTFL": "Y" if collected else "N", } ) os.makedirs(OUTPUT_DIR, exist_ok=True) with open( os.path.join(OUTPUT_DIR, "adsl.csv"), "w", newline="", encoding="utf-8", ) as handle: writer = csv.DictWriter(handle, fieldnames=FIELDNAMES) writer.writeheader() writer.writerows(adsl_rows)if __name__ == "__main__": main()solutions/adam-adsl-investigator-comment/result.sas
Edit101 lines/* opensas reference solution for the yamaa benchmark adam-adsl-investigator-comment. Reads the benchmark inputs and derives the requested output. Execute with the existing opensas runtime: opensas /app/output/result.sas. *//* Read DM line by line and split the fields here: a quoted comment can hold commas, doubled quotation marks, and line breaks, so each field's length is counted as it is read. An empty field, quoted or not, has length zero. */data dm; infile "/app/input/dm.csv" firstobs=2 truncover; length STUDYID $20 USUBJID $40 COMMENT $1000 FIELD $1000 CH $1; retain STUDYID USUBJID COMMENT FIELD "" CMNTLEN FLEN NFIELD INQUOTE 0; keep STUDYID USUBJID COMMENT CMNTLEN; input; if lengthn(_infile_) = 0 and not INQUOTE then delete; RECLEN = lengthc(_infile_); do I = 1 to RECLEN; CH = substr(_infile_, I, 1); if INQUOTE and CH = '"' then do; /* A doubled quotation mark inside quotes is one mark of the value. */ if I < RECLEN and substr(_infile_, I + 1, 1) = '"' then do; link addchar; I + 1; end; else INQUOTE = 0; end; else if not INQUOTE and CH = '"' then INQUOTE = 1; else if not INQUOTE and CH = "," then link endfield; else link addchar; end; /* A line break inside quotes is part of the comment; anywhere else it ends the record. */ if INQUOTE then do; CH = '0A'x; link addchar; end; else do; link endfield; output; call missing(STUDYID, USUBJID, COMMENT); CMNTLEN = 0; NFIELD = 0; end; return;/* Append one character, a space included, to the field being read. */addchar: if FLEN = 0 then FIELD = CH; else FIELD = substr(FIELD, 1, FLEN) || CH; FLEN + 1; return;/* Store the field just read under its column. */endfield: NFIELD + 1; if NFIELD = 1 then STUDYID = FIELD; else if NFIELD = 2 then USUBJID = FIELD; else if NFIELD = 3 then do; COMMENT = FIELD; CMNTLEN = FLEN; end; FIELD = ""; FLEN = 0; return;run;/* CMNT is the comment as recorded, CMNTFL Y when one was collected. A field of only spaces counts as no comment, so it has length zero too. */data dm; set dm; if lengthn(COMMENT) = 0 then CMNTLEN = 0;run;proc sql; create table adsl as select STUDYID, USUBJID, case when CMNTLEN > 0 then COMMENT else "" end as CMNT length=1000, case when CMNTLEN > 0 then "Y" else "N" end as CMNTFL length=1, CMNTLEN from dm order by STUDYID, USUBJID;quit;/* Write the file directly: the comment quoted, with its quotation marks doubled, at its recorded length. */data _null_; set adsl; file "/app/output/adsl.csv"; length QUOTED $2000 OUTLINE $2200; if _n_ = 1 then put "STUDYID,USUBJID,CMNT,CMNTFL"; if CMNTLEN = 0 then OUTLINE = trim(STUDYID) || "," || trim(USUBJID) || ",," || CMNTFL; else do; QUOTED = tranwrd(CMNT, '"', '""'); QLEN = CMNTLEN + countc(CMNT, '"'); OUTLINE = trim(STUDYID) || "," || trim(USUBJID) || ',"' || substr(QUOTED, 1, QLEN) || '",' || CMNTFL; end; put OUTLINE;run;
yamaa