Hewers of wood, namers of streets after trees

I was playing around with Where the Streets Have Your Name and became curious about the frequency of street names in Canada.

Using some open data sources, I was able to pull together a list of the most-common street names in Canada (long and tedious bit about the methodology appears below the table).

Oddly, “Second” is the most frequent street name in towns, villages and cities across the country. This seems counterintuitive until you realize that “First Street” is usually the main drag in a town and prone to renaming, often for a local hero or something bland imposed by provincial authorities.

Thus, what should be called First Street in New Glasgow, Nova Scotia, is now Mitchell Street. Instead of First Street in Denzil, Saskatchewan, there is Highway 676.

The remaining numbered streets follow Benford’s Law, so there are fewer Fifths than Fourths and Ninths than Eighths.

The top of the list also includes the blancmange Main and Park, of course, but a unique Canadian quirk is Railway so high on the list. This is due to the number of towns that sprung up along the Canadian Pacific Railway line, particularly on the Prairies.

Victoria is also a nice reference to our former Queen, in keeping with the era in which many of Canada’s smaller towns acquired proper street names.

And then there are the trees: Maple (and des Erables), Pine (des Pins), Birch, Spruce, Cedar, Willow, Poplar. Monty Python fans will be disappointed that “The Larch” does not make the top 200. There’s only 87 Larch streets.

Methodology: I thought pulling together this list would be a relatively simple task, but it was surprisingly complex. I couldn’t find any open-data sources with a full list of street names across the entire country. The best I could do was an enormous map file (an .shp) called the National Roads Network.
I imported this 250 MB file into ArcMap, a GIS program and found that while it includes what appears to be every street in the country, it doesn’t identify the community in which it is located. Worse, many streets list in the file are broken down into multiple segements, so Bank Street in Ottawa is listed multiple times and difficult to distinguish from Bank Place in Kamloops.

In order to sort this list, I needed to assign each street to its nearest town, village or city. Here, ArcMap was indispensable. I used a “spatial join” to calculate the distance of several million street segments to the nearest community. Even on a fast computer, this took a while.

Once each street segment was assigned its closest community, I exported the records to MySQL and used a GROUP BY to cluster the streets by community, then count the number of unique records you see listed above.

The approach in not perfect. Ideally, I would have had shapefiles with polygons for the boundaries of every community in the country. That’s not available for free, yet.

So, the number counts are probably not perfect, but close enough for fun. * Note: For the sake of simplicity, French and English versions of numbered streets were counted together, so that Second, 2nd, and 2e are grouped.