Proposed Analysis

[syounkin]

I would like to see the bikeable region divided into a grid. For each square/polygon of the grid we compute the bike route to the school of interest with brouter. For each route we compute a set of route-level measurements, e.g., distance, duration, % cycleway, average tls, etc. I noticed brouter provides some summary statistics in the front-end, e.g., Highway, Surface, Smoothness, etc. We could then report the proportion of students that don't have good/safe bike routes, where they live, and what makes it bad/unsafe. I'd like to give this a shot but it might be a little slower than our current pace. I'll have to sort through working with osrm/brouter and a new R class. Any pointers, sample code, etc. is appreciated. Thoughts?

[syounkin]

I would also like to consider defining the bikeable area using a minimum distance as well as a maximum distance. I think we should be excluding routes from the current cycling analysis that are less than half a mile long.

[bvarick]

I was thinking the same thing. Check out the new “routeslts” block for what I have so far. It calculates average and max lts for each route. I’m interested in a metric of “distance travelled at lts of 3 or higher”.

[bvarick]

I set eval=false on that block because it’s slow. There’s probably a way to make it more efficient