Source: I downloaded an [extract](https://download.geofabrik.de/north-america/us.html) of OpenStreetMap data for the US from Geofabrik. Then I used duckdb to extract all restaurants (sit down and fast-food) to a CSV file.
`COPY (`
`SELECT tags[‘name’][1] as name,`
`tags[‘brand:wikidata’][1] as brand_wikidata,`
`tags[‘cuisine’] as cuisine,`
`id`
`FROM st_readOSM(‘~/Downloads/us-latest.osm.pbf’)`
`WHERE (tags[‘amenity’]=[‘restaurant’] OR tags[‘amenity’] = [‘fast_food’])`
`AND tags[‘name’] != []`
`) TO ‘osm_us_restaurants.csv’;COPY (`
`SELECT tags[‘name’][1] as name,`
`tags[‘brand:wikidata’][1] as brand_wikidata,`
`tags[‘cuisine’] as cuisine,`
`id`
`FROM st_readOSM(‘~/Downloads/us-latest.osm.pbf’)`
`WHERE (tags[‘amenity’]=[‘restaurant’] OR tags[‘amenity’] = [‘fast_food’])`
`AND tags[‘name’] != []`
`) TO ‘osm_us_restaurants.csv’;`
Then I used pandas and seaborn to visualize the data as a heatmap. I used the following regex match to check for the frequency of each number.
numbers = pd.DataFrame({“number”: [str(num) for num in range (1, 101)]})
numbers[“us_restaurant_count”] = numbers.apply(lambda x: len(restaurants_without_chains[restaurants_without_chains.name.str.match(r”(^|s)” + x[“number”] + r”($|s)”, case=False)]), axis=1)numbers = pd.DataFrame({“number”: [str(num) for num in range (1, 101)]})
I feel like these are all just from Pho restaurants… Especially 88, which is lucky in Vietnamese (and Chinese).
I kinda expected some pink at 69.
Would have liked logarithmic scaling to see the distribution of uncommon numbers better.
I’m pleasantly surprised at the lack of 69 in restaurant names. Guess business owners are more mature than internet folk.
There was once a restaurant near me called “Thai 65,” which seems to have been more special than I gave it credit for.
As a non – US resident I expected a blip at 66 for motels and diners by the Route 66, and TIL it has mostly lost its meaning to the I-40.
My local Pho 20 cookin up that fire since 1998 🔥🔥🔥
I wonder what the logarithmic scale would look like. Might just be noise but I’m more interested in the weird and probably nonsensical connections of the handful of restaurants with X number in the title.
No offence OP, but this is pretty useless since 3 overpowers almost all other numbers. It should really be logarithmic or something so we can see the detail in the lower counts.
Not only is this representation not beautiful, this is also possibly one of the least interesting trends you could have picked.
Why 3? Tres Amigos?
I’d expect spikes at 66 (Route 66 diners) and 88 because of the good luck connotation.
15 comments
not liking the spike at 88
Well technically neither 1-9 nor 100 are 2-digit numbers, and there does not seem to be an option to solely contain the digit 0.
For more graphs like this, check out my latest [blog post](https://surprisedatespot.com/blog/trends-in-restaurant-names)!
Source: I downloaded an [extract](https://download.geofabrik.de/north-america/us.html) of OpenStreetMap data for the US from Geofabrik. Then I used duckdb to extract all restaurants (sit down and fast-food) to a CSV file.
`COPY (`
`SELECT tags[‘name’][1] as name,`
`tags[‘brand:wikidata’][1] as brand_wikidata,`
`tags[‘cuisine’] as cuisine,`
`id`
`FROM st_readOSM(‘~/Downloads/us-latest.osm.pbf’)`
`WHERE (tags[‘amenity’]=[‘restaurant’] OR tags[‘amenity’] = [‘fast_food’])`
`AND tags[‘name’] != []`
`) TO ‘osm_us_restaurants.csv’;COPY (`
`SELECT tags[‘name’][1] as name,`
`tags[‘brand:wikidata’][1] as brand_wikidata,`
`tags[‘cuisine’] as cuisine,`
`id`
`FROM st_readOSM(‘~/Downloads/us-latest.osm.pbf’)`
`WHERE (tags[‘amenity’]=[‘restaurant’] OR tags[‘amenity’] = [‘fast_food’])`
`AND tags[‘name’] != []`
`) TO ‘osm_us_restaurants.csv’;`
Then I used pandas and seaborn to visualize the data as a heatmap. I used the following regex match to check for the frequency of each number.
numbers = pd.DataFrame({“number”: [str(num) for num in range (1, 101)]})
numbers[“us_restaurant_count”] = numbers.apply(lambda x: len(restaurants_without_chains[restaurants_without_chains.name.str.match(r”(^|s)” + x[“number”] + r”($|s)”, case=False)]), axis=1)numbers = pd.DataFrame({“number”: [str(num) for num in range (1, 101)]})
numbers[“us_restaurant_count”] = numbers.apply(lambda x: len(restaurants_without_chains[restaurants_without_chains.name.str.match(r”(^|s)” + x[“number”] + r”($|s)”, case=False)]), axis=1)
Fuck Baskin Robbins 31 Flavors, I guess
I feel like these are all just from Pho restaurants… Especially 88, which is lucky in Vietnamese (and Chinese).
I kinda expected some pink at 69.
Would have liked logarithmic scaling to see the distribution of uncommon numbers better.
I’m pleasantly surprised at the lack of 69 in restaurant names. Guess business owners are more mature than internet folk.
There was once a restaurant near me called “Thai 65,” which seems to have been more special than I gave it credit for.
As a non – US resident I expected a blip at 66 for motels and diners by the Route 66, and TIL it has mostly lost its meaning to the I-40.
My local Pho 20 cookin up that fire since 1998 🔥🔥🔥
I wonder what the logarithmic scale would look like. Might just be noise but I’m more interested in the weird and probably nonsensical connections of the handful of restaurants with X number in the title.
No offence OP, but this is pretty useless since 3 overpowers almost all other numbers. It should really be logarithmic or something so we can see the detail in the lower counts.
Not only is this representation not beautiful, this is also possibly one of the least interesting trends you could have picked.
Why 3? Tres Amigos?
I’d expect spikes at 66 (Route 66 diners) and 88 because of the good luck connotation.
Does 1 include names like Ichiban?
Comments are closed.