Civic Knowledge

Eric Busboom.

San Diego’s CityIQ system consists of about 4,400 smart street lamps, which use a variety of sensors to track parking, pedestrians, traffic and a variety of environmental conditions. This analysis reveals pedestrian paterns using time-oriented heatmaps.

`` Full notebook on github

This dataset aggregates the number of pedestrians, traveling in either direction, into 15 minute intervals, per walkway, and links the walkways to communities. Then, data is vizualized by community in a heat map with two time dimensions, known as a rhythm map. Since this data is only available for the prior 6 months, we can’t analyze long term patterns, so we only present a few types of maps, including traffic for hour of day versus day of week and for hour of day versus week of year.

The US government has been widely criticized for its failure to test as many of its citizens for COVID-19 infections as other countries. But is mass testing really as easy as it seems? This analysis of the false positive and false negative rates of tests, using published sensitivities and specificities for COVID-19 rt-PCR and antigen tests, shows that even tests with slightly less than perfect results can produce very large numbers of false positives.

Eric Busboom.

The SIR infection model is a simple way of exploring the effects of a social distance policy on a large population such as San Diego County

`` Full notebook on github

The SIR infection model is a simple set of differential equations that describe how a disease can spread under very homogeneous, stable conditions. The model gets it’s name from the three “compartments” for people in diferent conditions:

Eric Busboom.

Exploring crime patterns with Ryhthm Maps, heat maps with two time dimensions.

`` Full notebook on github

A Rhythm map is a heat map with two time dimensions, which makes it much esier to find temporal patterns in some kinds of data. This example will use crime incidents from San Diego county, from 2007 to 2013..

Plotting the number of incidents for burglary versus the hour of the day, for instance, results in this dull chart:

Eric Busboom.

An isochrone is the area of travel from a point in a constant time, which allows for more realistic analysis of retail cachment areas than a simple radius.

An isochrone ( meaning “equal time” ) is an area that encloses the points one can travel to in a fixed time. For a bird, an isochrone would be a circle, but humans usually have to follow roads, so human isochrones have shapes that mirror the road network. Despite the name, isochrones are most often computed using a fixed distance. For instance, a 5km isochrone would be all the points that are within 5km of a point, but using distances based on a road network, not the straight distance. So, instead of a 5km radius circle, a 5km isochrone will have a non-circular shape.

Richard Florida’s article in the Wall Street Journal documents the early results of worker’s shift from offices to Zoom; falling rents and increasing vacancies in expensive large metros and the opposite trend in smaller less expensive metros, using data provided by Apartment List.

The Apartment List data shows a very broad decline in US rents compared to previous years, starting in with the COVID-19 pandemic in March of 2020. However, the national declines are primarily due to rent drops in expensive coastal cities. More affordable cities, however, have seen significant rent increases:

In preparation for the SCALE and Open San Diego Crime ANalysis Presentation, I downloaded the data package for the presentation and started on the analysis a bit early.

The presentation will use the San Diego Police 2015 calls for service, but the data repository has 2016 and 2017 as well. The seperate files for each of the three years are converted into a data package on our repository, and I also created Metapack packages for the boundary files for beats, neighborhoods and districts. Merging in these boundary files allows for analysing crime counts by geography. Here are the packages:

San Diego has the world’s largest municipal Internet-of-things deployment,  which I got access to recently and have been pulling data from. I’ve now got about 30 million parking records — events when a car enters or leaves a parking zone — and have been processing and analyzing the data for the last few weeks.

The raw data is a bit difficult to use, because there are a lot of spurious events, but it isn’t hard to clean them out. The result is records, at 15 minute intervals, of the number of cars parked in each of the 6,600 parking zones across the city. Here are the two processed datasets.

San Diego’s CityIQ system – the 4,400 smart street lamps installed across the city – yields detailed data about pedestrian movements. We’ve condensed a few hundred gigabytes of observations of pedestrians from August 2018 to June 2019 into a dataset of about 31M records, with the counts of the number of people walking in 2,600 walkways. I’d analyzed an earlier dataset to expose daily patterns of pedestrian traffic, and now, with a more complete dataset, I’m exploring other topics. The latest is a look at the streets where San Diego parties and works downtown.

The Los Angeles County Economic Roundtable maintains a data collection for the Los Angeles Point In Time homelessness counts, as well as data sets for the associated in person surveys. It’s a really remarkable collection, with information both about the locations of the homeless, as well as information about their demographics and health issues. The San Diego Regional Data Library has produced an analysis dataset of the survey data with a few re-coded variables, including Census geoids, that make analysis easier.