Skip to main content
100%

Massachusetts crime count by community

✓ Published1🌍 Public
Wwenleicao
Last edited Sep 18, 2019
Created on Sep 15, 2019

This visualization compares crime type distributions across Massachusetts communities using a small-multiple bar chart, with each panel representing a community type (city, suburban, or rural) and bars encoding logged crime counts for categories like burglary and property crime. The data is aggregated via `vl.aggregate` and grouped by community and crime type, then rendered with `vl.markBar`, using VEGA-Lite alongside D3 for CSV loading and Vega Tooltip for hover details. The logged y-axis reveals proportionally higher property crimes in rural areas despite lower absolute counts.

AI-generated description

A visualization constructed using the vega-lite-api.

Question related with this dataset

  1. geo distribution of crime in map?
  2. any crime type difference in city, suburban, and rural
  3. what are the top 5 crime/population in massachusetts?

I try to implement a visual to show 2.
the orignal dataset structure did not support, I have to unpivot I tried to use function of vega lite. but found there is a pivot function, but unpivot is not support yet. I end up unpivoting it in excel and load to the gist.

since rural area has very few record, while city has much more record, I have to use log scale to show rural. This might not a best options to show. Since the height may be distorted during to log scale. 100% stack column chart can show proportion better.

I forked Curran's normalized stack chart and modify a bit. see here.

This clearly indicated rural area has more burglary and property crime, probably due to few people.

dateset intro is here

Gist is from here

MIT Licensed

Similar vizzes