This dataset features a collection of R scripts designed to leverage the Affinity Propagation machine-learning algorithm for clustering similar window photographs. These scripts employ an unsupervised clustering analysis, culminating in the manual tagging of clusters. Following this, the scripts process the tags and clusters to generate tabulated data in .csv format. Additionally, the code includes tools for reapplying EXIF metadata to images and organizing them efficiently. The dataset also provides a sample output, derived from processing the scripts with data from the study “Image dataset: Year-long hourly facade photos of a university building” (https://doi.org/10.1016/j.dib.2024.110798).