BlockR: An Areal Spatial Anonymization and Visualization Tool

Citation to full paper: Haensch, Anna, and Claire Kelling. "BlockR: An Areal Spatial Anonymization and Visualization Tool." Transactions in GIS 29.4 (2025): e70070.

This repository contains the code necessary to implement the BlockR method to anonymize areal data.

First, we describe the helper files which includes functions to implement this method that we created.

• 00_bg_functions.R: This file contains the vast majority of the functions that we've written/adapted to support the BlockR method. This includes functions called blockR(), border_cells_fn(), clust_units(), connect_islands(), full_selb_eval_fun(), make_outlines(), make_outlines(), make_poly_new_coord(), outline_one_shp(), rotateProj(), round_poly_coords(), round_poly_coords_sub(), samp_prob(), selb_bound(), st_queen(), and st_rook(). We will add descriptions of these functions soon.
• 00_border_change.R: We created a function to randomly permute the border of the rotated and blockified shape, as described in our paper. This function is called border_change().

Next, we describe the helper files that were adapted from other packages, including some R packages that have been deprecated.

• 00_maptools_functions.R: Adaptation of maptools R package functions because some functions have been deprecated (including the elide() family of functions, rotateCoords(), scaleCoords(), and unionSpatialPolygons() adapted to our naming of unionSpatialPolygons_pck()).
• 00_surveillance_functions.R: Adaptation of surveillance R package function to detect shapes that are at the border of a polygon, (including polyAtBorder() adapted to our naming of polyAtBorder_cust()).
• 00_rgeos_functions.R: Adaptation of rgeos R package functions as some functions have been deprecated (including gUnaryUnion(), gBuffer_block(), rgeosStatus()).

Finally, we describe the files to implement the method, as is shown in the paper.

• 01_paper_figures.R: Reproduces the analysis in the paper shown in Figures 1, 4, 6, and 7.
• 02_blockR_reproduce.R: Reproduces the analysis in the paper shown in Figures 9, 10, 11.
• 03_images_for_classification.R: Creating images for the image classification analysis, summarized in Table 1. These include both random and non-random coloring of the blockified units.
• 04_blockR_illustration.R: The previous files are used to reproduce the analysis in our paper. If you are looking to utilize the BlockR method on a new place (e.g. county), we would suggest adapting this script for the applicable geographic area.

Code to implement our classification analysis (Table 1) is in a Google Colab notebook, which can be accessed by this link.

We welcome suggestions for improvements to our approach. If you utilize our method in your work, we would love to know. Please reach out to us at ckelling at carleton dot edu.