Free-Text Start Dates Read as ISO 8601 Dates

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

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

1# Reference solution for the yamaa benchmark adam-adcm-free-text-start-date (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)910cm <- read_csv(11  "/app/input/cm.csv",12  col_types = cols(CMSEQ = col_integer(), .default = col_character())13)1415months <- c("JAN", "FEB", "MAR", "APR", "MAY", "JUN",16            "JUL", "AUG", "SEP", "OCT", "NOV", "DEC")1718# The month a three-letter abbreviation names, in any letter case.19month_of <- function(abbreviation) match(toupper(abbreviation), months)2021# A date written as ISO 8601 text; an impossible date is an error.22iso_date <- function(year, month, day) {23  value <- as.Date(sprintf("%04d-%02d-%02d", as.integer(year), as.integer(month),24                           as.integer(day)), format = "%Y-%m-%d")25  if (is.na(value)) stop("not a date: ", year, "-", month, "-", day)26  format(value, "%Y-%m-%d")27}2829# The start as ISO 8601 text at the precision collected.30read_start <- function(text) {31  if (is.na(text) || text == "") return(NA_character_)32  if (grepl("^[0-9]{4}-[0-9]{2}-[0-9]{2}$", text)) {33    return(iso_date(substr(text, 1, 4), substr(text, 6, 7), substr(text, 9, 10)))34  }35  if (grepl("^[0-9]{2}[A-Za-z]{3}[0-9]{4}$", text) && !is.na(month_of(substr(text, 3, 5)))) {36    return(iso_date(substr(text, 6, 9), month_of(substr(text, 3, 5)), substr(text, 1, 2)))37  }38  if (grepl("^[0-9]{2}/[0-9]{2}/[0-9]{4}$", text)) {39    # The month comes first.40    return(iso_date(substr(text, 7, 10), substr(text, 1, 2), substr(text, 4, 5)))41  }42  if (grepl("^[0-9]{8}$", text)) {43    return(iso_date(substr(text, 1, 4), substr(text, 5, 6), substr(text, 7, 8)))44  }45  if (grepl("^[A-Za-z]{3}[0-9]{4}$", text) && !is.na(month_of(substr(text, 1, 3)))) {46    # A month alone.47    return(sprintf("%s-%02d", substr(text, 4, 7), month_of(substr(text, 1, 3))))48  }49  stop("CMSTDAT is in none of the five spellings: ", text)50}5152adcm <- cm |>53  mutate(54    ASTDTC = vapply(CMSTDAT, read_start, character(1), USE.NAMES = FALSE),55    # A month alone starts on its first day.56    ASTDT = if_else(nchar(ASTDTC) == 7, paste0(ASTDTC, "-01"), ASTDTC)57  ) |>58  select(STUDYID, USUBJID, CMSEQ, CMTRT, CMSTDAT, ASTDTC, ASTDT)5960dir.create("/app/output", showWarnings = FALSE)61write_csv(adcm, "/app/output/adcm.csv", na = "")

1# Reference solution for the yamaa benchmark adam-adcm-free-text-start-date (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"""Build ADCM from CM, reading a free-text start date in five spellings."""89import csv10import re11from datetime import date1213INPUT_DIR = "/app/input"14OUTPUT_DIR = "/app/output"1516MONTHS = [17    "JAN",18    "FEB",19    "MAR",20    "APR",21    "MAY",22    "JUN",23    "JUL",24    "AUG",25    "SEP",26    "OCT",27    "NOV",28    "DEC",29]303132def month_number(abbreviation):33    """The month a three-letter abbreviation names, in any letter case."""34    upper = abbreviation.upper()35    return MONTHS.index(upper) + 1 if upper in MONTHS else None363738def iso_date(year, month, day):39    return date(int(year), int(month), int(day)).isoformat()404142def read_start(text):43    """The start as ISO 8601 text at the precision collected, or ""."""44    if not text:45        return ""46    if re.fullmatch(r"[0-9]{4}-[0-9]{2}-[0-9]{2}", text):47        return iso_date(text[:4], text[5:7], text[8:])48    match = re.fullmatch(r"([0-9]{2})([A-Za-z]{3})([0-9]{4})", text)49    if match and month_number(match[2]):50        return iso_date(match[3], month_number(match[2]), match[1])51    match = re.fullmatch(r"([0-9]{2})/([0-9]{2})/([0-9]{4})", text)52    if match:53        # The month comes first.54        return iso_date(match[3], match[1], match[2])55    if re.fullmatch(r"[0-9]{8}", text):56        return iso_date(text[:4], text[4:6], text[6:])57    match = re.fullmatch(r"([A-Za-z]{3})([0-9]{4})", text)58    if match and month_number(match[1]):59        # A month alone.60        return f"{match[2]}-{month_number(match[1]):02d}"61    raise ValueError(f"CMSTDAT {text!r} is in none of the five spellings")626364def main():65    with open(f"{INPUT_DIR}/cm.csv", newline="") as handle:66        cm = list(csv.DictReader(handle))67    fieldnames = ["STUDYID", "USUBJID", "CMSEQ", "CMTRT", "CMSTDAT", "ASTDTC", "ASTDT"]68    with open(f"{OUTPUT_DIR}/adcm.csv", "w", newline="") as handle:69        writer = csv.DictWriter(handle, fieldnames=fieldnames)70        writer.writeheader()71        for row in cm:72            astdtc = read_start(row["CMSTDAT"])73            # A month alone starts on its first day.74            astdt = astdtc + "-01" if len(astdtc) == 7 else astdtc75            writer.writerow(76                {77                    "STUDYID": row["STUDYID"],78                    "USUBJID": row["USUBJID"],79                    "CMSEQ": row["CMSEQ"],80                    "CMTRT": row["CMTRT"],81                    "CMSTDAT": row["CMSTDAT"],82                    "ASTDTC": astdtc,83                    "ASTDT": astdt,84                }85            )868788if __name__ == "__main__":89    main()

1/* opensas reference solution for the yamaa benchmark adam-adcm-free-text-start-date.2   Reads the benchmark inputs and derives the requested output.3   Execute with the existing opensas runtime: opensas /app/output/result.sas. */45data cm;6  length STUDYID USUBJID CMTRT CMSTDAT $1024;7  infile "/app/input/cm.csv" dsd dlm="," firstobs=2 truncover lrecl=32767;8  input STUDYID : $char1024. USUBJID : $char1024. CMSEQ CMTRT : $char1024. CMSTDAT : $char1024.;9run;1011/* The start is read in five spellings: YYYY-MM-DD, DDMONYYYY and MONYYYY12   (the month in any letter case), MM/DD/YYYY with the month first, and13   YYYYMMDD. MONYYYY gives a month alone, which starts on its first day. */14data result;15  length ASTDTC $10 ASTDT $10 TEXT $1024;16  set cm;17  TEXT=upcase(strip(CMSTDAT));18  ASTDTC=''; ASTDT='';19  len=lengthn(TEXT);20  day=.;21  if len=10 and substr(TEXT, 5, 1)='-' then day=input(TEXT, yymmdd10.);22  else if len=9 then day=input(TEXT, date9.);23  else if len=10 and substr(TEXT, 3, 1)='/' then day=input(TEXT, mmddyy10.);24  else if len=8 then day=input(TEXT, yymmdd8.);25  else if len=7 then do;26    first=input(cats('01', TEXT), date9.);27    if not missing(first) then do;28      ASTDTC=substr(put(first, yymmdd10.), 1, 7);29      ASTDT=put(first, yymmdd10.);30    end;31  end;32  if not missing(day) then do;33    ASTDTC=put(day, yymmdd10.);34    ASTDT=ASTDTC;35  end;36run;3738proc sql;39  create table final as select STUDYID, USUBJID, CMSEQ, CMTRT, CMSTDAT, ASTDTC, ASTDT from result;40quit;41proc export data=final outfile="/app/output/adcm.csv" dbms=csv replace; run;