The locations of crime, fires, and emergency medical response calls often display distinctive spatial patterns in cities. These patterns are of interest for a variety of reasons, including allocating resources, planning, and identifying factors related to their occurrence. Many traditional statistical techniques are unable to address the spatial nature of these types of phenomena.

Geopublic has experience using a variety of methods that are designed specifically for the analysis of spatial data. Whether detecting spatial patterns in residential fires, or identifying factors related to neighbourhood crime rates, Geopublic has the necessary skills and tools.

Example 1. Detecting Clusters of Residential Fires

A variety of techniques are available for detecting significant spatial patterns in data. This example demonstrates a clustering technique, which we use to identify areas where residential fires are occurring closer together than would be expected if fires were distributed randomly throughout the city.

Example 2. Spatial Statistical Analysis of Crime

Spatial patterning in data can often violate certain assumptions made by traditional statistical techniques. Methods aimed at explaining neighbourhood crime rates in relation to other factors (e.g., poverty, etc.) must account for the underlying spatial patterns in the data.

This example depicts a special scatterplot that can be used to assess the degree of spatial patterning in the data. It indicates that the crime rate in one neighbourhood, is partially dependent on the crime rate in adjacent neighbourhoods. This dependence must be considered while attempting to identify factors that influence crime levels.