How to Come across an Apartment with a Shorter Commute | by Dávid Guszejnov | Mar, 2023

A guideline on how to make a custom made commuter’s map of your home city applying facts science and Google Maps API

Very last year I received a task in Massachusetts at the Centre for Astrophysics in Cambridge. When hunting for close by flats, 1 of my vital prerequisites was that it should not get too prolonged to commute to function. Due to the fact Boston has a fairly built up public transportation procedure (MBTA) with several subway traces, it was not clear in which neighborhoods I really should even be on the lookout for flats. So I decided to make a map of the Better Boston region demonstrating how very long it requires to commute to my workplace through morning rush hour. This publish information how I built the map and how you can do one particular for yourself. You can locate the source on GitHub.

Warning: If you want to do a equivalent venture (i.e., a map for a distinctive spot/town) you have to get your possess API key. Google Maps enables you to do a moderate selection of calls for no cost for each month, so it is achievable to do a job like this for no cost. Even so, you should be extremely watchful, it is straightforward to go in excess of the restrict and then be charged hundreds of pounds.

Calculating Commute Moments

The first action was to get a map of the Boston area and make a mapping in between it and GPS coordinates (i.e., latitude and longitude for just about every pixel). On this map we can determine a grid of details the place we will calculate the vacation time. To reduce the quantity of API phone calls I chose to make my grid factors extra dense in the vicinity of the middle and significantly less dense on the outskirts. Take note that Google Maps instantly “snaps” to close by streets, so we really don’t will need to fear about no matter whether our grid details are really on a highway.

Once we have the grid, we just need to contact Google Maps and use its Distance Matrix API to calculate the journey moments to our desired destination. Nevertheless, there are a few subtleties to hold in thoughts when calculating vacation instances:

1. We need to have to specify the time of the day. Due to the fact I would be commuting to my workplace during the early morning hurry hour I established an arrival time of 9 AM.
2. Google Maps can offer vacation occasions for driving, biking and taking general public transportation. For my task I picked general public transportation only.
3. Most varieties of general public transportation are fairly rare in the US, even for the duration of rush hour (e.g., buses coming only each and every 15 minutes). This can introduce an error in our journey time calculation, so to minimize its effects I made the decision to work out the travel situations for 2 far more arrival times (8:45, 8:52) and take the minimal of the three values. This effectively suggests that I would be ready to occur into do the job a little bit sooner if that suggests not waiting at the bus cease for 20 minutes.

Commute Time overlaid on Town Map

The moment we have the commute times for each grid issue, we can visualize them on the town map with a crammed contour plot.

As envisioned, commute time boosts with length, but we can also notice anomalies, factors physically even further becoming nearer by commute time. There are even embedded islands of decreased commute times. This is due to the structure of the MBTA network. For instance, living around the Kendall/MIT subway station we can get to our location in 30 min, but if we lived several streets nearer to our place we would need to consider a bus and would get there later on. There are also smaller islands from in which it is not achievable to reach our vacation spot (e.g., coach restore middle in East Cambridge).

Commute Length of Boston Neighborhoods

When this map is handy, it would be improved to have a thing we could use to filter final results from apartment listing websites. Most of these sites checklist which neighborhood each condominium belongs to, making it possible for us to filter for it. So, it would make feeling for us to translate our commute time map into a map of neighborhoods. To start with, let us just make a map of the Boston neighborhoods.

We could check out to attract a map of the Boston neighborhoods the place the distance of any stage to my office (the new origin) is proportional to the commute time (alternatively of the physical distance). We can carry out this by shifting the length of every pixel relative of the origin to be proportional to the commute time, though preserving their relative direction the exact.

phi = np.arctan2(y_impression,x_image)
x_new = commute_time * np.cos(phi)
y_new = commute_time * np.sin(phi)

This will distort the picture, and direct to uneven pixel dimensions (i.e., some pixels will be crowding each and every other, some will have gaps involving them). We can correct for that by doing a Voronoi diagram and coloring the ensuing cells according the the colour of the corresponding pixel.

from scipy.spatial import Voronoi, voronoi_plot_2d
vor = Voronoi(np.vstack((x_new,y_new)).T)
voronoi_plot_2d(vor,ax=ax,clearly show_points=Wrong,clearly show_vertices=Phony,line_width=.)
...
#Colorize the Voronoi plot
for i,location in enumerate(vor.areas):
coloration = ...
polygon = [vor.vertices[k] for k in location]
plt.fill(*zip(*polygon),c=color)

With this map we can immediately see which neighborhoods are near sufficient for us to commute from. Take note that some places in adjacent neighborhoods get combined (e.g., Downtown place). This is owing to the presence of superior velocity community transportation (e.g., subway), which tends to make it quicker to commute from a station 1 stop further more than from an condominium that is a 5 minute walk from the present station.

This job commenced with me hunting for an condominium from which my commute would not be also extended(< 45 min). This means I should mainly be looking for apartments in Cambridge, Somerville, Belmont, Arlington, Allston, Watertown and the Downtown area. Of course there are other considerations when buying an apartment (rent, noise etc.). In the end I rented an apartment near the boundary of Cambridge and Arlington.

If you want to try to do the same for your city, check out the source on GitHub.