Input: a CSV district,month,pm25, a line ---, then a caption (sentences ending with . or !).
Check each sentence with these rules (first matching rule):
<District> rose|fell by N%— compare with (last − first)/first of that district (months in order): supported if the direction matches and |actual| is within 1 percentage point of N; printSUPPORTED (actual -23.1%): <sentence>orNOT SUPPORTED (actual -5.3%): …;<District> had the highest average…— supported if the district has the highest mean:… (highest is Lucheng at 46.3);The city average was N/city average was N— supported if the mean of all rows is within 0.5:… (actual 41.7);- otherwise
UNCHECKED: <sentence> (no rule for this kind of claim - needs a human).
Finish with 3 of 4 checkable claims supported.