1 of 18

Options for conducting near repeat analysis

@routineactivity

Image: Metropolitan Police

2 of 18

Near Repeats using ArcGIS Pro

3 of 18

4 of 18

Repeat and Near Repeat Classification

Inputs

Point Data (X/Y Coords)
Date of incident
Repeat Incident Distance (0 = same address or building)
Spatial Bands (meters)
Temporal Bands (days)

5 of 18

6 of 18

Near Repeats using JR NR Calculator and open-source GIS (QGIS)

7 of 18

Download Jerry Ratcliffe Near Repeat Calculator v2.0 and Manual here

8 of 18

Download Jerry Ratcliffe Near Repeat Calculator and Manual here

9 of 18

10 of 18

11 of 18

Dates and sequences need to be applied manually, but there is a graphical modeller and Python script within QGIS that would allow this to be automated. The lines added are within PowerPoint, but within the Visualist Plugin in QGIS there are options including ‘Flow map’ that could be used, again to save time.

**There is a Near Repeat package and a Near Strings script which exist in Python that could be scripted into pyQGIS and run as an automatic function, similar to the ArcPro GIS out of the box tool. This is worth the time investment if you wanted to run near repeats daily/weekly for operational or tactical analysis**

Whilst there is a QGIS Cloud/WebMap it’s unlikely something that could be used within a law enforcement agency. Outputs therefore would likely need to be provided the old fashioned way, as static images accompanied by other analysis or information in briefing or tasking slides, or BOLO bulletins.

12 of 18

Near Repeats using JR NR Calculator output with BI dashboards (Microsoft PowerBI)

13 of 18

An alternative output for the near repeats output from the JR NR calculator is a Business Intelligence dashboard.

This could allow some reproducibility and automation. Once the output is run from JR NR calc (not tried, but may be possible to save time here using PowerAutomate/PowerApps with Microsoft), a refresh of a BI dashboard like the one shown is done with one mouse click. Where agencies/organisations are using PowerBI or alternatives, end users will be able to view and interrogate the outputs interactively and choose to view Originators/Near Repeats and time parameters.

In this example using the Icon Map custom visual in PowerBI (https://icon-map.com/), this produces a Leaflet map with Open Street Map base map. This is customisable, tooltips and labelling can also be added.

This can become almost entirely automated/reproducible.

14 of 18

Near Repeats using R NearRepeat and ptools

15 of 18

NearRepeat & ptools packages

Walked-through example can be found here:

https://rpubs.com/routineactivity/crime_near_repeats

16 of 18

Near Repeats using Python

17 of 18

Python NearRepeat

https://github.com/tobydavies/NearRepeat

Python Near Strings https://andrewpwheeler.com/2017/04/12/identifying-near-repeat-crime-strings-in-r-or-python/

18 of 18

Slides: link��Contact: @routineactivity

@routineactivity

Image: Metropolitan Police