The Battle of Neighborhoods - Top 10 Restaurants in 8 Boroughs of Toronto by Rating and Likes

Introduction

Toronto is the capital city of the Canadian province of Ontario. Its current area is 630.20 𝑘𝑚2 (243.3 square mile). With a recorded population of 2,731,571 in 2016, it is the most populous city in Canada and the fourth most populous city in North America. Toronto is a prominent centre for music, theatre, motion picture production, and television production, and is home to the headquarters of Canada's major national broadcast networks and media outlets. Its varied cultural institutions, which include numerous museums and galleries, festivals and public events, entertainment districts, national historic sites, and sports activities, attract over 43 million tourists each year.

If you're one of those 43 million tourists, you might have many concerns when you visit Toronto. These concerns may include place to go, eat, sleep, shop and so on.

This project aims to select top 10 restaurants in 8 boroughs(Downtown Toronto, West Toronto, East Toronto, Central Toronto, North York, Scarborough, East York, Etobicoke) of Toronto based on number of likes and rating restaurants from Foursquare. Foursquare is a local search-and-discovery mobile app developed by Foursquare Labs Inc. It also has web app and API.

Data

Data To discover restaurants of concern, following data are used:

Foursquare API : Forsquare API to get the most common restaurants of a given neighbourhood of a borough. It's in json format.
Postal Code of Boroughs in Toronto : Postal code of neighbourhoods of a given borough in Toronto. This data can be obtained from wikipedia using Pandas. This data consist of Postal Code, Borough and Neighbourhood. It's in Tabular format.
Neighbourhood Latitude and Longitude : Neighbourhood Latitude and Longitude of a given borough. Latitude and Longitude are used to obtain the most common restaurants in a given neighbourhood using Foursquare API. It can be found in Geospatial_Coordinates.csv. It's a csv file

Methodology

The target is to retrieve top 10 restaurants in 8 boroughs of Toronto. That are Downtown Toronto, West Toronto, East Toronto, Central Toronto, North York, Scarborough, East York and Etobicoke.

Technology Used

Numpy: Scientific Computing libray for Python
Pandas: Data manipulation and analysis
Matplotlib: Visualization
requests: API operations
Nominatim: Convert location name to Latitude and Longitude
folium: Visualize data on map

Data Extraction

First, I get postal codes in 8 boroughs of Toronto from wikipedia page using Pandas, and filter out not assigned column value. I use Pandas to read geospatial goordinates of toronto from Geospatial_Coordinates.csv. Then I join the postal codes with geospatial goordinates.

Here is what we have so far: toronto nb latlong

After that, I use Foursquare API to explore venues in each neighbourhood using latitude and longitude with the following endpoind: https://api.foursquare.com/v2/venues/explore?&client_id={}&client_secret={}&v={}&ll={},{}&radius={}&limit={}.

This is what we have get: toronto venues

So now, let's filter the data to keep Venue Category that are restaurants.
toronto retaurants

Next, let's use Venue ID for each restaurants to get their details with the following endpoind:
https://api.foursquare.com/v2/venues/{}?&client_id={}&client_secret={}&v={}.
I choose following column: Venue ID, Address, Phone Number, Likes, Rating, and Tips.

restaurants' detail

We have restaurants' detail. Let's merge restaurants with their detail, base on Venue ID, to get information we need for restaurants in each neighbourhood.
toronto Retaurants Detail

Finally, sort all restaurants in descending order and group them base on borough and take top 10 of them. top 10 restaurants

Results

There is one thing that is needed to point out. From 10 top restaurants above, there are just 65 restaurants. We have 8 boroughs, and we are interested top 10 restaurants in each borough. This means that, base on Foursquare API, some boroughs have less than 10 restaurants. When visualize results, I decide to call restaurants in boroughs with 10 or more restaurants top 10 restaurants. Beside this, let's just call them top restaurants.

I start with bar chart. After bar chart, I create a visualization for all top/top 10 restaurants on the map using Name, number of like and rating of a given restaurant as a pop-up.

Downtown Toronto

Downton Toronto Top10

West Toronto

Chicago Geo plot

East Toronto

East Town top10

Central Toronto

Central Toronto top10

North York

North York top10

Scarborough

East York

Etobicoke

All top/top 10 restaurants on the map

Discussion

The purpose of this project is to find Top 10 Restauarants in 8 Boroughs of Toronto by Rating and Likes using Foursquare API. For some boroughs, there are less than 10 restaurants. This is a drawback.

Toronto itself varies in culture. It attracts millions of people every year. Further analysis should be made. It would be interesting to use different API to explore more.

Base on Foursquare API, although it is not complete, this might give us a pretty good idea of restaurants in Toronto and its boroughs. The reason is the data is based on people.

Conclusion

Starting by introducing of problem, then describe needed data, walkthrough methodology, get the results, and finally the discussion of whether the project is helpful. I hope this gives help target audiens in deciding of choosing restaurants in each boroughs and Toronto overall.

Future Work

Try different API
Discover top restaurants in different category (Italian, Japanese...)
find out, overall, which borough has better restaurants (Likes and Rating)
Examine tip(review) from user