Rank comparables by weighted similarity.

The similarity engine from CValR, ported faithfully: min–max normalisation over subject + comps, then a weighted Gower distance to your subject, ranked. Load a mapped CSV, pick the subject, set the weights.