Crime Analysis

 Crime Analysis

GIS 5100 Applications in GIS - Module 1

This lab covered three types of hot spot mapping:

• Grid Based Thematic Mapping
• Kernel Density
• Local Moran's I

Each of the processes involved is discussed and pictured below: 

Grid Based Thematic Mapping:

First the environmental parameters were updated so that the extent and map were for layer Boundary_Chicago. The grid and homicide point layers were joined and then select by attribute was used to create a new layer with only the records that reported 1 or more homicides. To select the top 20% of the records, the attribute table was sorted in descending order and the top 214 rows were selected (this number was determined by dividing the total number of records by 5 and rounding down to the nearest whole number. In order to convert the map at its current stage to a polygon, the add new field, calculate field and dissolve tools were used. 

Kernel Density:

The first step in this process is to run the geoprocessing Kernel Density tool using the input features provided in the lab. This uses the 2017 homicide shape file to produce the output raster ( cell size 110ft, search radius 2630ft) showing densities. The show statistics option under symbology panel > Classes tab>more>show statistics was used to determine the mean. This information was then used to calculate the class ranges ( in this case only 2 classes: 3x mean and greater than 3x mean). IN order to get a polygon that has only the higher of the two classes, it was first reclassified and then converted using the raster to polygon tool, then a selection of the desired class (value 2) was exported to produce the final output feature class. 

Local Moran's I: 

To prepare the data for the Moran’s I tool, the census tracts and 2017 homicide features were spatially joined. Then the crime rate was calculated by adding a field and using the calculate field tool with a formula that using the count divided by the number of households multiplied by 1000. Now the data is processed using the Cluster and Outlier Analysis tool. This results in a number of values. For this lab we are only interested in the “HH” or High-High values. These values are selected and made into a new feature class. This feature class is then dissolved into a single shape.