This dataset contains 250 points that correspond to buildings where cholera deaths were recorded near the Broad Street pump. These data were shared publicly by Wilson (2011). The variables include an ID for each building, its coordinates, and a death count. The following variables were added: an ID for the closest pump, as well as the distance to it and to the Broad Street pump (in meters), and, finally, a dummy variable that classifies observations in terms of whether the Broad Street pump was the closest pump to them.

The documentation is available at:


The Geoda script is available at:


Overview of data

Variable Description
ID Unique ID
deaths Number of deaths per building
pumpID ID of the nearest pump (see ‘pumps’ dataset)
distpump Distance to the nearest pump (in meters - see ‘pumps’ dataset)
distBSpump Distance to Broad St pump (in meters - see ‘pumps’ dataset)
BSpump Create categories depending on whether the Broad Street pump is closest (1) or not (0)
COORD_X X-coordinate in meters (see documentation for projection details).
COORD_Y Y-coordinate in meters (see documentation for projection details).


Wilson, R (2011). John Snow’s Cholera data in more formats. Reprojected data can also be downloaded from Dani Arribas-Bel’s ‘reproducible john snow’ BitBucket repository at

Prepared by Center for Spatial Data Science. Last updated February 1, 2021. Data provided “as is,” no warranties.