Fig. 4From: Cleaning by clustering: methodology for addressing data quality issues in biomedical metadataTwo clusters for the key “age”, see panels (a) and (b), respectivelyBack to article page