Build One Record per Collected Lab Result

1library(yamaa)23lb <- yamaa_domain("spec.yaml")$output4lb

1import yamaa23lb = yamaa.yamaa_domain("spec.yaml").output4lb

1# Reference solution for the yamaa benchmark sdtm-lb-findings (R track).2#3# Written from the benchmark's full prompt and its inputs alone, as an agent4# would write it. Harbor's oracle agent runs it to check that the prompt can5# be solved and that the grader scores a correct answer 1.67library(dplyr, warn.conflicts = FALSE)8library(readr)9library(tidyr)1011odm <- read_csv(12  "/app/input/odm.csv",13  col_types = cols(.default = col_character())14)1516key_cols <- c(17  "StudyOID", "SubjectKey", "StudyEventOID", "StudyEventRepeatKey",18  "ItemGroupOID", "ItemGroupRepeatKey"19)2021dates <- odm |>22  filter(ItemOID == "IT.LB.LBDTC") |>23  select(all_of(key_cols), LBDTC = Value)2425lb <- odm |>26  filter(ItemOID %in% c("IT.LB.CALCIUM", "IT.LB.CREAT")) |>27  left_join(dates, by = key_cols) |>28  mutate(Value = na_if(Value, "")) |>29  filter(!is.na(Value)) |>30  mutate(31    LBTESTCD = if_else(ItemOID == "IT.LB.CALCIUM", "CA", "CREAT"),32    LBTEST = if_else(ItemOID == "IT.LB.CALCIUM", "Calcium", "Creatinine"),33    NOTDONE = Value == "NOT DONE",34    DOMAIN = "LB",35    STUDYID = StudyOID,36    USUBJID = SubjectKey,37    LBORRES = if_else(NOTDONE, NA_character_, Value),38    LBORRESU = if_else(NOTDONE, NA_character_, "mg/dL"),39    LBSTRESN = if_else(40      NOTDONE, NA_real_, suppressWarnings(as.numeric(Value))41    ),42    LBSTRESU = if_else(NOTDONE, NA_character_, "mg/dL"),43    LBSTAT = if_else(NOTDONE, "NOT DONE", NA_character_)44  ) |>45  arrange(STUDYID, USUBJID, LBDTC, LBTESTCD) |>46  group_by(STUDYID, USUBJID) |>47  mutate(LBSEQ = row_number()) |>48  ungroup() |>49  select(50    DOMAIN, STUDYID, USUBJID, LBSEQ, LBTESTCD, LBTEST, LBORRES,51    LBORRESU, LBSTRESN, LBSTRESU, LBSTAT, LBDTC52  )5354dir.create("/app/output", showWarnings = FALSE)55write_csv(lb, "/app/output/lb.csv", na = "")

1# Reference solution for the yamaa benchmark sdtm-lb-findings (Python track).2#3# Written from the benchmark's full prompt and its inputs alone, as an agent4# would write it. Harbor's oracle agent runs it to check that the prompt can5# be solved and that the grader scores a correct answer 1.67from pathlib import Path89import polars as pl1011odm = pl.read_csv("/app/input/odm.csv", infer_schema=False)1213key_cols = [14    "StudyOID",15    "SubjectKey",16    "StudyEventOID",17    "StudyEventRepeatKey",18    "ItemGroupOID",19    "ItemGroupRepeatKey",20]2122wide = (23    odm.group_by(key_cols)24    .agg(pl.struct("ItemOID", "Value").alias("items"))25    .with_columns(26        pl.col("items")27        .list.eval(28            pl.element().filter(29                pl.element().struct.field("ItemOID") == "IT.LB.LBDTC"30            )31        )32        .list.get(0)33        .struct.field("Value")34        .alias("LBDTC"),35    )36    .drop("items")37)3839tests = odm.filter(pl.col("ItemOID").is_in(["IT.LB.CALCIUM", "IT.LB.CREAT"]))4041code_of = {"IT.LB.CALCIUM": "CA", "IT.LB.CREAT": "CREAT"}42name_of = {"CA": "Calcium", "CREAT": "Creatinine"}4344joined = tests.join(wide, on=key_cols, how="left")454647def as_number(text):48    if text is None or text == "" or text == "NOT DONE":49        return None50    try:51        return float(text)52    except ValueError:53        return None545556rows = []57for r in joined.to_dicts():58    value = r["Value"]59    # A test with no collected entry produces no record.60    if value is None or value == "":61        continue62    code = code_of[r["ItemOID"]]63    not_done = value == "NOT DONE"64    rows.append(65        {66            "DOMAIN": "LB",67            "STUDYID": r["StudyOID"],68            "USUBJID": r["SubjectKey"],69            "LBDTC": r["LBDTC"],70            "LBTESTCD": code,71            "LBTEST": name_of[code],72            "LBORRES": None if not_done else value,73            "LBORRESU": None if not_done else "mg/dL",74            "LBSTRESN": None if not_done else as_number(value),75            "LBSTRESU": None if not_done else "mg/dL",76            "LBSTAT": "NOT DONE" if not_done else None,77        }78    )7980lb = pl.DataFrame(81    rows,82    schema={83        "DOMAIN": pl.String,84        "STUDYID": pl.String,85        "USUBJID": pl.String,86        "LBDTC": pl.String,87        "LBTESTCD": pl.String,88        "LBTEST": pl.String,89        "LBORRES": pl.String,90        "LBORRESU": pl.String,91        "LBSTRESN": pl.Float64,92        "LBSTRESU": pl.String,93        "LBSTAT": pl.String,94    },95).sort(["STUDYID", "USUBJID", "LBDTC", "LBTESTCD"]).with_columns(96    pl.int_range(1, pl.len() + 1).over(["STUDYID", "USUBJID"]).alias("LBSEQ")97).select(98    "DOMAIN",99    "STUDYID",100    "USUBJID",101    "LBSEQ",102    "LBTESTCD",103    "LBTEST",104    "LBORRES",105    "LBORRESU",106    "LBSTRESN",107    "LBSTRESU",108    "LBSTAT",109    "LBDTC",110)111112Path("/app/output").mkdir(parents=True, exist_ok=True)113lb.write_csv("/app/output/lb.csv")

1/* opensas reference solution for the yamaa benchmark sdtm-lb-findings.2   Reads the benchmark inputs and derives the requested output.3   Execute with the existing opensas runtime: opensas /app/output/result.sas. */45data odm;6  length StudyOID MetaDataVersionOID SubjectKey StudyEventOID StudyEventRepeatKey ItemGroupOID ItemGroupRepeatKey ItemOID Value $1024;7  infile "/app/input/odm.csv" dsd dlm="," firstobs=2 truncover lrecl=32767;8  input StudyOID : $char1024. MetaDataVersionOID : $char1024. SubjectKey : $char1024. StudyEventOID : $char1024. StudyEventRepeatKey : $char1024. ItemGroupOID : $char1024. ItemGroupRepeatKey : $char1024. ItemOID : $char1024. Value : $char1024.;9run;1011proc sort data=odm out=ordered; by StudyOID SubjectKey StudyEventOID StudyEventRepeatKey ItemGroupRepeatKey; run;12data dates;13  length LBDTC $1024;14  retain LBDTC;15  set ordered;1617  by StudyOID SubjectKey StudyEventOID StudyEventRepeatKey ItemGroupRepeatKey;18  if first.ItemGroupRepeatKey then do;19    LBDTC='';20  end;21    if ItemOID='IT.LB.LBDTC' then LBDTC=Value;22  if last.ItemGroupRepeatKey then output;23run;24proc sql;25create table joined as select a.*,b.LBDTC from odm a inner join dates b on a.StudyOID=b.StudyOID and a.SubjectKey=b.SubjectKey and a.StudyEventOID=b.StudyEventOID and a.StudyEventRepeatKey=b.StudyEventRepeatKey and a.ItemGroupRepeatKey=b.ItemGroupRepeatKey where a.ItemOID in ('IT.LB.CALCIUM','IT.LB.CREAT') and not missing(a.Value); quit;26data prepared; set joined; length DOMAIN $2 STUDYID USUBJID LBTESTCD LBTEST LBORRES LBORRESU LBSTRESU LBSTAT $1024;27DOMAIN='LB'; STUDYID=StudyOID; USUBJID=SubjectKey; LBORRES=Value; LBORRESU='mg/dL'; LBSTRESU='mg/dL'; LBSTRESN=.; LBSTAT='';28if Value ne 'NOT DONE' then LBSTRESN=input(Value,?? best32.);29if ItemOID='IT.LB.CREAT' then do; LBTESTCD='CREAT'; LBTEST='Creatinine'; end; else do; LBTESTCD='CA'; LBTEST='Calcium'; end;30if Value='NOT DONE' then do; LBORRES=''; LBORRESU=''; LBSTRESU=''; LBSTRESN=.; LBSTAT='NOT DONE'; end; run;31proc sort data=prepared; by STUDYID USUBJID LBDTC LBTESTCD; run;32data result; set prepared; by STUDYID USUBJID; retain LBSEQ;33  if first.USUBJID then LBSEQ=0;34  LBSEQ+1;35run;3637proc sql;38  create table final as select DOMAIN, STUDYID, USUBJID, LBSEQ, LBTESTCD, LBTEST, LBORRES, LBORRESU, LBSTRESN, LBSTRESU, LBSTAT, LBDTC from result;39quit;40proc export data=final outfile="/app/output/lb.csv" dbms=csv replace; run;