Time-Matched ECG Baselines by Period

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

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

1# Reference solution for the yamaa benchmark adam-adeg-time-matched-baseline (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)910adeg <- read_csv(11  "/app/input/adeg_raw.csv",12  col_types = cols(13    APERIOD = col_integer(), APERDAY = col_integer(),14    ATPTN = col_double(), AVAL = col_double(),15    .default = col_character()16  ),17  na = ""18)1920# Day -1 of each period is that period's baseline day.21baseline <- adeg |>22  filter(APERDAY == -1) |>23  select(STUDYID, USUBJID, PARAMCD, APERIOD, ATPTN, BASE = AVAL)2425# Time-matched: the same subject, parameter, period and nominal time on the26# period's own baseline day; nothing stands in for a missing one.27adeg_out <- adeg |>28  left_join(29    baseline,30    by = c("STUDYID", "USUBJID", "PARAMCD", "APERIOD", "ATPTN"),31    relationship = "many-to-one"32  ) |>33  mutate(34    ABLFL = if_else(APERDAY == -1, "Y", NA_character_),35    CHG = if_else(APERDAY == -1, NA_real_, AVAL - BASE),36    BASETYPE = paste0("Period ", APERIOD, " Baseline Day ", ATPT)37  ) |>38  select(39    STUDYID, USUBJID, PARAMCD, PARAM, APERIOD, APERDAY, ATPT, ATPTN, AVAL,40    AVALU, ABLFL, BASE, CHG, BASETYPE41  )4243dir.create("/app/output", showWarnings = FALSE)44write_csv(adeg_out, "/app/output/adeg.csv", na = "")

1# Reference solution for the yamaa benchmark adam-adeg-time-matched-baseline (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 1.67"""Create ADEG from ADEG_RAW with time-matched period baselines.89For each subject, parameter and analysis period, the baseline day is10APERDAY -1. BASE is the AVAL recorded at the same nominal time point on11that period's baseline day; CHG is AVAL - BASE (blank on the baseline12day itself and wherever no baseline value exists).13"""1415import csv16from decimal import Decimal1718INPUT_PATH = "/app/input/adeg_raw.csv"19OUTPUT_PATH = "/app/output/adeg.csv"2021FIELDS = [22    "STUDYID", "USUBJID", "PARAMCD", "PARAM", "APERIOD", "APERDAY",23    "ATPT", "ATPTN", "AVAL", "AVALU", "ABLFL", "BASE", "CHG", "BASETYPE",24]252627def main():28    with open(INPUT_PATH, newline="") as f:29        rows = list(csv.DictReader(f))3031    # AVAL on each period's baseline day, by subject/parameter/period/time point.32    baseline = {}33    for row in rows:34        if row["APERDAY"] == "-1":35            key = (row["USUBJID"], row["PARAMCD"], row["APERIOD"], row["ATPT"])36            baseline[key] = row["AVAL"]3738    out_rows = []39    for row in rows:40        is_baseline_day = row["APERDAY"] == "-1"41        key = (row["USUBJID"], row["PARAMCD"], row["APERIOD"], row["ATPT"])42        base = baseline.get(key, "")43        if is_baseline_day or base == "":44            chg = ""45        else:46            chg = str(Decimal(row["AVAL"]) - Decimal(base))47        out = {name: row[name] for name in FIELDS[:10]}48        out["ABLFL"] = "Y" if is_baseline_day else ""49        out["BASE"] = base50        out["CHG"] = chg51        out["BASETYPE"] = (52            "Period " + row["APERIOD"] + " Baseline Day " + row["ATPT"]53        )54        out_rows.append(out)5556    with open(OUTPUT_PATH, "w", newline="") as f:57        writer = csv.DictWriter(f, fieldnames=FIELDS)58        writer.writeheader()59        writer.writerows(out_rows)606162if __name__ == "__main__":63    main()

1/* opensas reference solution for the yamaa benchmark adam-adeg-time-matched-baseline.2   Reads the benchmark inputs and derives the requested output.3   Execute with the existing opensas runtime: opensas /app/output/result.sas. */45data adeg;6  length STUDYID USUBJID PARAMCD PARAM ATPT AVALU $1024;7  infile "/app/input/adeg_raw.csv" dsd dlm="," firstobs=2 truncover lrecl=32767;8  input STUDYID : $char1024. USUBJID : $char1024. PARAMCD : $char1024. PARAM : $char1024. APERIOD APERDAY ATPT : $char1024. ATPTN AVAL AVALU : $char1024.;9run;1011/* Day -1 of each period is that period's baseline day; its records are the12   time-matched baselines, one per subject, parameter, period and time. */13data baseline;14  length B_STUDYID B_USUBJID B_PARAMCD $1024;15  set adeg;16  if APERDAY = -1;17  B_STUDYID = STUDYID; B_USUBJID = USUBJID; B_PARAMCD = PARAMCD;18  B_APERIOD = APERIOD; B_ATPTN = ATPTN; BASE = AVAL;19  keep B_STUDYID B_USUBJID B_PARAMCD B_APERIOD B_ATPTN BASE;20run;2122proc sql;23  create table matched as24    select a.*, b.BASE25    from adeg as a left join baseline as b26      on a.STUDYID = b.B_STUDYID and a.USUBJID = b.B_USUBJID27      and a.PARAMCD = b.B_PARAMCD and a.APERIOD = b.B_APERIOD28      and a.ATPTN = b.B_ATPTN;29quit;3031/* No change on the baseline day, and none without a baseline. */32data final;33  length ABLFL $1 BASETYPE $1024;34  set matched;35  if APERDAY = -1 then ABLFL = 'Y';36  else ABLFL = '';37  if APERDAY ne -1 and not missing(BASE) and not missing(AVAL) then CHG = AVAL - BASE;38  else CHG = .;39  BASETYPE = 'Period ' || strip(put(APERIOD, 8.)) || ' Baseline Day ' || strip(ATPT);40run;4142proc sql;43  create table out as44    select STUDYID, USUBJID, PARAMCD, PARAM, APERIOD, APERDAY, ATPT, ATPTN,45           AVAL, AVALU, ABLFL, BASE, CHG, BASETYPE46    from final;47quit;48proc export data=out outfile="/app/output/adeg.csv" dbms=csv replace; run;