Flag Measurable Disease at Baseline

1library(yamaa)23adrs <- yamaa_domain("spec.yaml")$output4adrs

1import yamaa23adrs = yamaa.yamaa_domain("spec.yaml").output4adrs

1# Reference solution for the yamaa benchmark adam-adrs-measurable-disease (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)910adsl <- read_csv(11  "/app/input/adsl.csv",12  col_types = cols(.default = col_character())13)14tu <- read_csv(15  "/app/input/tu.csv",16  col_types = cols(TUSEQ = col_integer(), .default = col_character())17)1819# Target disease at screening: a screening tumor identification record20# whose result is target. Later target records do not count.21target_subjects <- tu |>22  filter(23    TUTESTCD %in% "TUMIDENT",24    VISIT %in% "SCREENING",25    TUSTRESC %in% "TARGET"26  ) |>27  distinct(USUBJID)2829adrs <- adsl |>30  mutate(31    PARAMCD = "MDIS",32    PARAM = "Measurable Disease at Baseline",33    AVALC = if_else(USUBJID %in% target_subjects$USUBJID, "Y", "N"),34    AVAL = if_else(AVALC == "Y", 1L, 0L)35  ) |>36  select(STUDYID, USUBJID, PARAMCD, PARAM, AVALC, AVAL) |>37  arrange(USUBJID)3839dir.create("/app/output", showWarnings = FALSE)40write_csv(adrs, "/app/output/adrs.csv", na = "")

1# Reference solution for the yamaa benchmark adam-adrs-measurable-disease (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 pl1011adsl = pl.read_csv("/app/input/adsl.csv", infer_schema=False)12tu = pl.read_csv("/app/input/tu.csv", infer_schema=False).with_columns(13    pl.col("TUSEQ").cast(pl.Int64)14)1516# Target disease at screening: a screening tumor identification record17# whose result is target. Later target records do not count.18target_subjects = tu.filter(19    (pl.col("TUTESTCD") == "TUMIDENT")20    & (pl.col("VISIT") == "SCREENING")21    & (pl.col("TUSTRESC") == "TARGET")22).select("USUBJID").unique()2324adrs = (25    adsl.with_columns(26        PARAMCD=pl.lit("MDIS"),27        PARAM=pl.lit("Measurable Disease at Baseline"),28        AVALC=pl.when(pl.col("USUBJID").is_in(target_subjects["USUBJID"]))29        .then(pl.lit("Y"))30        .otherwise(pl.lit("N")),31    )32    .with_columns(33        AVAL=pl.when(pl.col("AVALC") == "Y").then(1).otherwise(0).cast(pl.Int64)34    )35    .select("STUDYID", "USUBJID", "PARAMCD", "PARAM", "AVALC", "AVAL")36    .sort("USUBJID")37)3839Path("/app/output").mkdir(exist_ok=True)40adrs.write_csv("/app/output/adrs.csv")

1/* opensas reference solution for the yamaa benchmark adam-adrs-measurable-disease.2   Reads the benchmark inputs and derives the requested output.3   Execute with the existing opensas runtime: opensas /app/output/result.sas. */45data adsl;6  length STUDYID USUBJID $1024;7  infile "/app/input/adsl.csv" dsd dlm="," firstobs=2 truncover lrecl=32767;8  input STUDYID : $char1024. USUBJID : $char1024.;9run;1011data tu;12  length STUDYID USUBJID TUTESTCD VISIT TUSTRESC $1024;13  infile "/app/input/tu.csv" dsd dlm="," firstobs=2 truncover lrecl=32767;14  input STUDYID : $char1024. USUBJID : $char1024. TUSEQ TUTESTCD : $char1024. VISIT : $char1024. TUSTRESC : $char1024.;15run;1617proc sql;18  create table counts as select STUDYID,USUBJID,count(*) as matches from tu where TUTESTCD='TUMIDENT' and VISIT='SCREENING' and TUSTRESC='TARGET' group by STUDYID,USUBJID;19  create table joined as select a.*,b.matches from adsl as a left join counts as b on a.STUDYID=b.STUDYID and a.USUBJID=b.USUBJID;20quit;21data result;22  length PARAMCD $8 PARAM $40 AVALC $1;23  set joined; PARAMCD='MDIS'; PARAM='Measurable Disease at Baseline';24  AVALC='N'; AVAL=0; if matches>0 then do; AVALC='Y'; AVAL=1; end;25run;2627proc sql;28  create table final as select STUDYID, USUBJID, PARAMCD, PARAM, AVALC, AVAL from result;29quit;30proc export data=final outfile="/app/output/adrs.csv" dbms=csv replace; run;