<p>Fuzzy partition clustering is a key unsupervised method that has attracted significant research interest. To better reveal the intrinsic relationships among data categories, researchers have recently introduced harmonic fuzzy partition clustering. However, its slow convergence speed limits its effectiveness for large-scale data processing. This paper proposes a modification strategy for harmonic fuzzy membership and develops a rapid harmonic fuzzy partition C-means clustering algorithm. Using the affinity filtering technique, we efficiently identify non-affinity clustering centers while minimizing computational costs. We then adjust the harmonic fuzzy membership degrees, enhancing those for affinity centers and resetting them for non-affinity centers. Our rapid harmonic fuzzy C-means algorithm significantly reduces convergence time by an average of 75% in iterations while maintaining excellent clustering performance and improving computational efficiency. This work advances harmonic fuzzy clustering theory and provides new support for data analysis and processing.</p> Graphical abstract <p></p>