<result> <column_name>Df%</column_name> <family>multiplicative</family> <subtype>ratio_multiplicative</subtype> <confidence>0.5</confidence> <requires_user_confirmation>true</requires_user_confirmation> <sum_meaningful>false</sum_meaningful> <signed>false</signed> <reducers> <primary_reducer>product</primary_reducer> <valid_reducers>product, geomean, min, max</valid_reducers> <excluded_reducers>sum (meaningless for ratios), arithmetic_mean (wrong average for ratios)</excluded_reducers> <average_semantics>geometric_mean</average_semantics> </reducers> <inner_outer_measure> <inner_measure_reducer>min</inner_measure_reducer> <outer_measure_reducer>max</outer_measure_reducer> </inner_outer_measure> <multiplicative_partner_derivable>false</multiplicative_partner_derivable> <reasoning>The table is PATH with a Date sequence axis, so multiplicative is plausible. The "%" in the name and the percent-formatted sample values (0.0%, 0.5%, 0.6%) point to a rate or relative change, which is the ratio_multiplicative pattern. The column is VARCHAR with placeholder values ("-", blank), so there are no numeric statistics, no center-of-mass or log-symmetry evidence, and no cumprod reconstruction to confirm it. The meaning of "Df" is ambiguous, and the column could be a share or level rather than a compounding rate. Confidence is therefore low and the call needs human confirmation. The values must be parsed to numbers, with placeholders treated as null, before any reduction. There is no evidence of a cross-product of incompatible measures, so the subtype is not tensorial.</reasoning> <table_type_conflict>false</table_type_conflict> </result>