Dose Intervals

1library(yamaa)23ex <- yamaa_domain("spec.yaml")$output4ex

1import yamaa23ex = yamaa.yamaa_domain("spec.yaml").output4ex

1# Reference solution for the yamaa benchmark sdtm-ex-intervals (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)910ec_raw <- read_csv(11  "/app/input/ec_raw.csv",12  col_types = cols(.default = col_character())13)1415ex <- ec_raw |>16  mutate(17    EXDOSE_NUM = suppressWarnings(as.numeric(ECDOSE)),18    ECSTDTC_DT = suppressWarnings(as.Date(ECSTDTC)),19    ECADJ_CLEAN = na_if(ECADJ, "")20  ) |>21  group_by(STUDYID, USUBJID, ECTRT, EXDOSE_NUM, ECDOSU, ECDOSFRQ) |>22  summarise(23    START_DT = min(ECSTDTC_DT, na.rm = TRUE),24    END_DT = max(ECSTDTC_DT, na.rm = TRUE),25    EXADJ = {26      vals <- sort(unique(ECADJ_CLEAN[!is.na(ECADJ_CLEAN)]))27      if (length(vals) == 0) NA_character_ else vals[1]28    },29    .groups = "drop"30  ) |>31  mutate(32    DOMAIN = "EX",33    EXTRT = ECTRT,34    EXDOSE = EXDOSE_NUM,35    EXDOSU = ECDOSU,36    EXDOSFRQ = ECDOSFRQ,37    EXSTDTC = START_DT,38    EXENDTC = END_DT39  ) |>40  arrange(STUDYID, USUBJID, START_DT, EXTRT) |>41  group_by(STUDYID, USUBJID) |>42  mutate(EXSEQ = row_number()) |>43  ungroup() |>44  select(45    DOMAIN, STUDYID, USUBJID, EXSEQ, EXTRT, EXDOSE, EXDOSU,46    EXDOSFRQ, EXSTDTC, EXENDTC, EXADJ47  )4849dir.create("/app/output", showWarnings = FALSE)50write_csv(ex, "/app/output/ex.csv", na = "")

1# Reference solution for the yamaa benchmark sdtm-ex-intervals (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 pl1011ec_raw = pl.read_csv("/app/input/ec_raw.csv", infer_schema=False)1213ec = ec_raw.with_columns(14    EXDOSE_NUM=pl.col("ECDOSE").cast(pl.Float64, strict=False),15    ECSTDTC_DT=pl.col("ECSTDTC").str.to_date(strict=False),16    ECADJ_CLEAN=pl.col("ECADJ").map_elements(17        lambda v: None if v is None or v == "" else v, return_dtype=pl.String18    ),19)2021grouped = (22    ec.group_by("STUDYID", "USUBJID", "ECTRT", "EXDOSE_NUM", "ECDOSU", "ECDOSFRQ")23    .agg(24        pl.col("ECSTDTC_DT").min().alias("START_DT"),25        pl.col("ECSTDTC_DT").max().alias("END_DT"),26        pl.col("ECADJ_CLEAN").drop_nulls().unique().sort().first().alias("EXADJ"),27        pl.col("ECDOSE").first().alias("EXDOSE_RAW"),28    )29    .with_columns(30        DOMAIN=pl.lit("EX"),31        EXTRT=pl.col("ECTRT"),32        EXDOSE=pl.col("EXDOSE_NUM"),33        EXDOSU=pl.col("ECDOSU"),34        EXDOSFRQ=pl.col("ECDOSFRQ"),35        EXSTDTC=pl.col("START_DT"),36        EXENDTC=pl.col("END_DT"),37    )38    .sort(["STUDYID", "USUBJID", "START_DT", "EXTRT"])39    .with_columns(EXSEQ=pl.col("USUBJID").cum_count().over(["STUDYID", "USUBJID"]))40    .with_columns(EXSEQ=pl.col("EXSEQ").cast(pl.Int64))41    .select(42        "DOMAIN",43        "STUDYID",44        "USUBJID",45        "EXSEQ",46        "EXTRT",47        "EXDOSE",48        "EXDOSU",49        "EXDOSFRQ",50        "EXSTDTC",51        "EXENDTC",52        "EXADJ",53    )54)5556Path("/app/output").mkdir(exist_ok=True)57grouped.write_csv("/app/output/ex.csv")

1/* opensas reference solution for the yamaa benchmark sdtm-ex-intervals.2   Reads the benchmark inputs and derives the requested output.3   Execute with the existing opensas runtime: opensas /app/output/result.sas. */45data ec_raw;6  length STUDYID USUBJID ECTRT ECDOSU ECDOSFRQ ECSTDTC ECADJ $1024;7  infile "/app/input/ec_raw.csv" dsd dlm="," firstobs=2 truncover lrecl=32767;8  input STUDYID : $char1024. USUBJID : $char1024. ECTRT : $char1024. ECDOSE ECDOSU : $char1024. ECDOSFRQ : $char1024. ECSTDTC : $char1024. ECADJ : $char1024.;9run;1011proc sql;12create table intervals as select STUDYID,USUBJID,ECTRT as EXTRT,ECDOSE as EXDOSE,ECDOSU as EXDOSU,ECDOSFRQ as EXDOSFRQ,min(ECSTDTC) as EXSTDTC,max(ECSTDTC) as EXENDTC from ec_raw group by STUDYID,USUBJID,ECTRT,ECDOSE,ECDOSU,ECDOSFRQ;13create table reasons as select STUDYID,USUBJID,ECTRT,ECDOSE,ECDOSU,ECDOSFRQ,min(ECADJ) as EXADJ from ec_raw where not missing(ECADJ) group by STUDYID,USUBJID,ECTRT,ECDOSE,ECDOSU,ECDOSFRQ;14create table prepared as select 'EX' as DOMAIN,a.*,b.EXADJ from intervals a left join reasons b on a.STUDYID=b.STUDYID and a.USUBJID=b.USUBJID and a.EXTRT=b.ECTRT and a.EXDOSE=b.ECDOSE and a.EXDOSU=b.ECDOSU and a.EXDOSFRQ=b.ECDOSFRQ; quit;15proc sort data=prepared; by STUDYID USUBJID EXSTDTC EXTRT; run;16data result; set prepared; by STUDYID USUBJID; retain EXSEQ;17  if first.USUBJID then EXSEQ=0;18  EXSEQ+1;19run;2021proc sql;22  create table final as select DOMAIN, STUDYID, USUBJID, EXSEQ, EXTRT, EXDOSE, EXDOSU, EXDOSFRQ, EXSTDTC, EXENDTC, EXADJ from result;23quit;24proc export data=final outfile="/app/output/ex.csv" dbms=csv replace; run;