Converting Point Data to Kernel Density Layer and to Polygon Data

preview_player
Показать описание
For Jedi Geographers!

This video tutorial demonstrates the multi-step process by which you can turn multiple years of point data (representing individual crimes) into a graduated point layer (using the collect events tool) and then into a kernel density 'hot spot' map (raster layer) and then convert the values from the raster layer's pixels (kernel density) into a table that can be joined to a polygon map (in this case a census block group layer) - so that you can get a count (or rate) of crimes in a neighborhood that would reduce the problem of crimes counting only in a single census tract (zip code, or other polygons based map) when crimes should perhaps be accounted to some degree in adjacent polygons, zip codes or census tracts as well. This also helps mitigate randomly slightly misplaced boundaries lines, or inaccurate geocoding that tends to create overcounts and/or undercounts crimes on one side of a polygonal boundary.

Data: Los Angeles Sheriff Department (2021)

Music:
Music by Complete Strangers, Song "Sweet Disarray"
Рекомендации по теме
Комментарии
Автор

THANK I needed this for my dissertation and I did not know how to get past the Kernel Density tool. Thank you so much for posting this video!

mattchewkimbo
Автор

Do you know if its possible to run KDE for polygon layers? In my case I don't have point data, but the area of study is very big (all municipalities in Mexico). Hence, I would like to show a smoother visualization of highway accidents with a raster instead of a cloropleth map. Thanks in advance in case you know any method or solution to this problem.

guillermocastillokoschnick