Module 1 Crime Analysis

For Module 1 Crime Analysis we covered Hotspot analysis creating three Hotspot maps in the Chicago region as I have shared here below. We were also working in the Washington DC region for the first portion of the lab working with exploring crime analysis. Here are some of the learning outcomes from this module, becoming familiar with techniques in crime analysis, examine spatial patterns in crime rates and socio-economic characteristics and compare the reliability of crime hotspot mapping for crime prediction.

For the First Map we used Grid-Based Thematic Mapping technique. I used the Spatial Join tool to create a new feature class with a newly added field for the homicide count per cell between the grid cells and the 2017 homicides. Then used Select By Attribute to select all grids with a homicide count greater than zero to create a new feature class. Then using quintile classification to take the top 20% of the grid cells with the highest count. Finishing up with using the Dissolve tool to create a multipart feature with a smooth service.

For the Second Map we used the Kernel Density Mapping technique. Using the Kernel Density tool I create a new feature class then edited the symbology to show only two breaks 3*[mean] and the maximum value. Then I used the Reclassify tool after being done reclassifying I used the Raster to Polygon tool to create a new feature. Then I used Select by Attributes to select the gridcode value of 2, then creating a feature class from the selection.

For the Third Map we used the Local Moran's I technique. I used the Spatial Join tool to create a new feature with the copied data of the census to the 2017 homicides adding a new field for the counts. Then I added a new field named crime rate to the newly create feature and used the field calculator to determine the number of homicides per 1000 units. Then I used the Cluster and Outlier Analysis (Moran's I) Spatial Statistic tool to create a new feature class to find the different level of crime rate from low to high-high. Then I used a SQL query to select the high-high clusters creating a feature class from the selection. Then finished with using the Dissolve tool, dissolving by the COType IDW field.