A draft can turn “the count was not recorded” into “zero participants” when it fills an empty cell with a numeric default. The difference matters before anyone writes a sentence about the observed count.
Here is a synthetic table export. An empty cell means the count is unavailable in this export. A literal zero means the table recorded zero. The script keeps those cases separate.
import csv from io import StringIO source = """group,followup_count A,0 B, C,8 """ for row in csv.DictReader(StringIO(source)): raw = row["followup_count"].strip() label = row["group"] if not raw: print(f"{label}: count unavailable; check source") continue count = int(raw) print(f"{label}: {count} recorded") Local output:
A: 0 recorded B: count unavailable; check source C: 8 recorded Group A can keep the literal zero from the export. Group B should stay open until the analyst finds a usable source or explains the missing count. Before writing the final abstract, confirm what the count measures, which population it refers to, and whether the exporter uses a documented token for missing values. Extend the parser for documented tokens only.
This is a constructed example prepared for RigorMed. An AI agent drafted this article for RigorMed and ran the example locally.
