
Vancouver Crime Analysis
METHOD
CHOROPLETH & HOTSPOT
Part 1: Excel
-
Sort and filter the Crime CSV obtained from the Vancouver Police Department to include only crime data from 2016
Part 2: ArcGIS
-
Create a geodatabase using the Neighbourhoods and CT shapefiles.
-
Clip the CT shapefile using the Neighbourhoods layer as the output feature. This new layer (CT_Clip) will be used later on to perform the hotspot analysis.
-
Import the 2016 Crime data using the Add XY data function and export it as a layer file.
-
Join the 2016 Crime data table to the Neighbourhoods layer. Added a field (Crime Type) and assigned each type of crime a numerical value.
-
Select by Attributes using an SQL Query to extract the winter (December, January, February) and summer (June, July, August) months from the 2016 Crime table.
-
Use the Collect Events tool from the Spatial Statistics toolbox on 2016 Crime layer to create a feature class containing weighted points. These weighted points are calculated by summing all incidents at each unique location. Repeat this step for the Winter and Summer Crime layers.
-
Perform an Optimized Hotspot Analysis using the 2016 collect events output layer. Count incidents analyzed within fishnet polygons using the CT_Clip layer as the bounding polygon. Repeat for Winter and Summer Crime layers. For choropleths count incidents were analyzed within aggregate polygons
-
Create a density surface using the Geostatistical Analyst Wizard using the newly created Hotspot layers.
REGRESSION & GROUPING ANALYSIS
Part 3​: Data
-
Obtain 2016 dwelling data from Statistics Canada, CHASS Centre
-
Sort and filter the dwelling data
-
Download Dissemination Area Layer Statistics Canada, Abacus Data Network
Part 4: Regression Analysis
-
Join data to a dissemination area base map
-
Join to a local area map
-
Join crime data to the joined DA layer
-
Spatial statistics tool -> modelling spatial relationships -> ordinary least squares
Part 5: Grouping Analysis
-
Spatial Statistics Tool -> Mapping Clusters -> Grouping Analysis
-
Using the results we were able to classify the different groups
DATA
For crime data, type of crime, location, and time from Geodash open data
https://geodash.vpd.ca/opendata/
Income from Statistics Canada, University of Toronto CHASS Centre
http://datacentre.chass.utoronto.ca/census
DA/CT boundaries from Statistics Canada, Abacus Dataverse Network
Neighbourhoods, Streets, City Boundary from City of Vancouver Open Data Catalogue
http://data.vancouver.ca/datacatalogue/index.htm
The Vancouver Police Department
https://vancouver.ca/police/organization/planning-research-audit/district-statistics.html
PROJECTION
NAD_1983_UTM_Zone_10N