The Battle of Neighborhoods

Puneet Arora
12 min readJun 6, 2021


Clustering Neighborhoods of London and Paris using Machine Learning


London and Paris are quite popular tourist and vacation destinations for people all around the world. They are diverse and multicultural and offer a wide variety of experiences that are widely sought after. We try to group the neighborhoods of London and Paris respectively and draw insights into what they look like now.

Business Problem

The aim is to help tourists choose their destinations depending on the experiences that the neighborhoods have to offer and what they would want to have. This model also helps people make decisions if they are thinking about migrating to London or Paris or even if they wish to relocate neighborhoods within the city. Our findings will help stakeholders make informed decisions and address any concerns they have, including the different kinds of cuisines, provision stores, and what the city has to offer.

Data Description

We require geographical location data for both London and Paris. Postal codes in each city serve as a starting point. Using Postal codes, we can find out the neighbourhoods, boroughs, venues and their most popular venue categories.


To derive our solution, We scrape our data from

This Wikipedia page has information about all the neighborhoods; we limit it to London.

  1. borough: Name of Neighborhood
  2. town: Name of the borough
  3. post_code: Postal codes for London.

This Wikipedia page lacks information about geographical locations. To solve this problem, we use ArcGIS API.


ArcGIS Online enables you to connect people, locations, and data using interactive maps. Work with smart, data-driven styles and intuitive analysis tools that deliver location intelligence. Share your insights with the world or specific groups.

More specifically, we use ArcGIS to get the geographical locations of the neighborhoods of London. Adding the following columns to our initial data set prepares our data.

  1. latitude: Latitude for Neighborhood
  2. longitude: Longitude for Neighborhood


To derive our solution, We leverage JSON data available at

The JSON file has data about all the neighborhoods in France; we limit it to Paris.

  1. postal_code: Postal codes for France
  2. nom_comm: Name of Neighborhoods in France
  3. nom_dept: Name of the boroughs, equivalent to towns in France
  4. geo_point_2d: Tuple containing the latitude and longitude of the Neighborhoods.

Foursquare API Data

We will need data about different venues in different neighborhoods of that specific borough. To gain that information, we will use “Foursquare” location information. Foursquare is a location data provider with information about all manner of venues and events within an area of interest. Such information includes venue names, locations, menus, and even photos. As such, the foursquare location platform will be used as the sole data source since all the stated required information can be obtained through the API.

After finding the list of neighborhoods, we then connect to the Foursquare API to gather information about venues inside each neighborhood. For each neighborhood, we have chosen the radius to be 500 meters.

The data retrieved from Foursquare contained information of venues within a specified distance of the longitude and latitude of the postcodes. The information obtained per venue as follows:

  1. Neighborhood: Name of the Neighborhood
  2. Neighborhood Latitude: Latitude of the Neighborhood
  3. Neighborhood Longitude: Longitude of the Neighborhood
  4. Venue: Name of the Venue
  5. Venue Latitude: Latitude of Venue
  6. Venue Longitude: Longitude of Venue
  7. Venue Category: Category of Venue

Based on all the information collected for both London and Paris, we have sufficient data to build our model. We cluster the neighborhoods together based on similar venue categories. We then present our observations and findings. Using this data, our stakeholders can take the necessary decision.


We will be creating our model with the help of Python, so we start by importing all the required packages. The code is available on GitHub to follow along.

import pandas as pd
import requests
import numpy as np
import as cm
import matplotlib.colors as colors
import folium
from sklearn.cluster import KMeans

Package breakdown:

  • Pandas: To collect and manipulate data in JSON and HTML and then data analysis
  • requests: Handle HTTP requests
  • matplotlib: Detailing the generated maps
  • folium: Generating maps of London and Paris
  • sklearn: To import K Means machine learning model.

The approach taken here is to explore each of the cities individually, plot the map to show the considered neighborhoods, and then build our model by clustering all of the similar neighborhoods together and finally plot the new map with the clustered neighborhoods. We draw insights and then compare and discuss our findings.

Data Collection

In the data collection stage, we begin with collecting the required data for the cities of London and Paris. We need data that has the postal codes, neighborhoods, and boroughs specific to each of the cities.

To collect data for London, using pandas, we scrape the List of areas of London Wikipedia page to take the 2nd table:

url_london = ""
wiki_london_url = requests.get(url_london)
wiki_london_data = pd.read_html(wiki_london_url.text)
wiki_london_data = wiki_london_data[1]
  1. nom_comm: Name of Neighborhoods in France
  2. nom_dept: Name of the boroughs, equivalent to towns in France
  3. geo_point_2d: Tuple containing the latitude and longitude of the Neighborhoods.

The data looks like this:

Data for London scrapped from the Wikipedia page.

To collect data for Paris, we download the JSON file containing all the postal codes of France from

Using Pandas, we load the table after reading the JSON file:

!wget -q -O 'france-data.json'
print("Data Downloaded!")
paris_raw = pd.read_json('france-data.json')

JSON data containing postal codes of France

Data Preprocessing

For London, We replace the spaces with underscores in the title. The borough column has numbers within square brackets that we remove using:

wiki_london_data.rename(columns=lambda x: x.strip().replace(" ", "_"), inplace=True)
wiki_london_data['borough'] = wiki_london_data['borough'].map(lambda x: x.rstrip(']').rstrip('0123456789').rstrip('['))

For Paris, we break down each of the nested fields and create the dataframe that we need:

paris_field_data = pd.DataFrame()
for f in paris_raw.fields:
dict_new = f
paris_field_data = paris_field_data.append(dict_new, ignore_index=True)


Feature Selection

For both of our datasets, we need only the borough, neighborhood, postal codes, and geolocations (latitude and longitude). So we end up selecting the columns that we need by:

df1 = wiki_london_data.drop( [ wiki_london_data.columns[0], wiki_london_data.columns[4], wiki_london_data.columns[5] ], axis=1)

df_2 = paris_field_data[['postal_code','nom_comm','nom_dept','geo_point_2d']]

Feature Engineering

Both of our Datasets contain information related to all the cities in the country. We can narrow down and further process the data by selecting only the neighborhoods of ‘London’ and ‘Paris’.

df1 = df1[df1['town'].str.contains('LONDON')]

df_paris = df_2[df_2['nom_dept'].str.contains('PARIS')].reset_index(drop=True)

Looking over our London dataset, we can see that we don’t have the geolocation data. We need to extrapolate the missing data for our neighborhoods. We perform this by leveraging the ArcGIS API. With the Help of, ArcGIS API we can get the latitude and longitude of our London neighborhood data.

from arcgis.geocoding import geocode
from arcgis.gis import GIS
gis = GIS()

Defining London ArcGIS geocode function to return latitude and longitude.

def get_x_y_uk(address1):
lat_coords = 0
lng_coords = 0
g = geocode(address='{}, London, England, GBR'.format(address1))[0]
lng_coords = g['location']['x']
lat_coords = g['location']['y']
return str(lat_coords) +","+ str(lng_coords)

Passing postal codes of London to get the geographical coordinates

coordinates_latlng_uk = geo_coordinates_uk.apply(lambda x: get_x_y_uk(x))

We proceed with Merging our source data with the geographical co-ordinates to make our dataset ready for the next stage

london_merged = pd.concat([df1,lat_uk.astype(float), lng_uk.astype(float)], axis=1)
london_merged.columns= ['borough','town','post_code','latitude','longitude']

The resulting feature engineered London Dataset

As for our Paris dataset, we don’t need to get the Geo coordinates using an external data source or collect them with the ArcGIS API call since we already have it stored in the geo_point_2d column as a tuple in the df_paris dataframe.

We just need to extract the latitude and longitude from the column:

paris_lat = paris_latlng.apply(lambda x: x.split(',')[0])
paris_lat = paris_lat.apply(lambda x: x.lstrip('['))

paris_lng = paris_latlng.apply(lambda x: x.split(',')[1])
paris_lng = paris_lng.apply(lambda x: x.rstrip(']'))

paris_geo_lat = pd.DataFrame(paris_lat.astype(float))

paris_geo_lng = pd.DataFrame(paris_lng.astype(float))

We then create our Paris dataset with the required information:

paris_combined_data = pd.concat([df_paris.drop('geo_point_2d', axis=1), paris_geo_lat, paris_geo_lng], axis=1)

Resulting Feature engineered dataset of Paris

Note: Both the datasets have been properly processed and formatted. Since the same steps are applied to both the datasets of London and Paris, we will be discussing the code for only the London dataset for simplicity.

Visualizing the Neighborhoods of London and Paris

Now that our datasets are ready, using the Folium package, we can visualize the maps of London and Paris with the neighborhoods that we collected.

Map of London showing the neighborhoods:

Map of Paris showing the neighborhoods:

Now that we have visualized the neighborhoods, we need to find out what each neighbourhood is like and what are the common venue and venue categories within a 500 m radius.

This is where Foursquare comes into play. With the help of Foursquare we define a function which collects information pertaining to each neighborhood including that of the name of the neighbourhood, Geo-coordinates, venue, and venue categories.


def getNearbyVenues(names, latitudes, longitudes, radius=500):

for name, lat, lng in zip(names, latitudes, longitudes):

# create the API request URL
url = '{}&client_secret={}&v={}&ll={},{}&radius={}&limit={}'.format(

# make the GET request
results = requests.get(url).json()["response"]['groups'][0]['items']

# return only relevant information for each nearby venue
v['venue']['categories'][0]['name']) for v in results])

nearby_venues = pd.DataFrame([item for venue_list in venues_list for item in venue_list])
nearby_venues.columns = ['Neighbourhood',
'Neighbourhood Latitude',
'Neighbourhood Longitude',
'Venue Category']


Data collected using Foursquare API

One Hot Encoding

Since we are trying to find out what are the different kinds of venue categories present in each neighborhood and then calculate the top 10 common venues to base our similarity on, we use the One Hot Encoding to work with our categorical data type of the venue categories. This helps to convert the categorical data into numeric data.

We won’t be using label encoding in this situation since label encoding might cause our machine learning model to have a bias or a sort of ranking which we are trying to avoid by using One Hot Encoding.

We perform one-hot encoding and then calculate the mean of the grouped venue categories for each of the neighborhoods.

# One hot encoding
London_venue_cat = pd.get_dummies(venues_in_London[['Venue Category']], prefix="", prefix_sep="")

# Adding neighbourhood to the mix
London_venue_cat['Neighbourhood'] = venues_in_London['Neighbourhood']

# moving neighborhood column to the first column
fixed_columns = [London_venue_cat.columns[-1]] + list(London_venue_cat.columns[:-1])
London_venue_cat = London_venue_cat[fixed_columns]

# Grouping and calculating the mean
London_grouped = London_venue_cat.groupby('Neighbourhood').mean().reset_index()

Data with the mean values for each of the Neighborhoods after one-hot encoding

Top Venues in the Neighborhoods

In our next step, We need to sort, rank, and label the top venue categories in our neighborhood.

Let’s define a function to get the top venue categories in the neighborhood

def return_most_common_venues(row, num_top_venues):
row_categories = row.iloc[1:]
row_categories_sorted = row_categories.sort_values(ascending=False)

return row_categories_sorted.index.values[0:num_top_venues]

There are many categories, we will consider the top 10 categories to avoid data skew.

Defining a function to label them accurately

num_top_venues = 10

indicators = ['st', 'nd', 'rd']

# create columns according to number of top venues
columns = ['Neighbourhood']
for ind in np.arange(num_top_venues):
columns.append('{}{} Most Common Venue'.format(ind+1, indicators[ind]))
columns.append('{}th Most Common Venue'.format(ind+1))

Getting the top venue categories in the neighborhoods of London

# create a new dataframe for London
neighborhoods_venues_sorted_london = pd.DataFrame(columns=columns)
neighborhoods_venues_sorted_london['Neighbourhood'] = London_grouped['Neighbourhood']

for ind in np.arange(London_grouped.shape[0]):
neighborhoods_venues_sorted_london.iloc[ind, 1:] = return_most_common_venues(London_grouped.iloc[ind, :], num_top_venues)


Top Common venues in each Neighborhood ranked.

Model Building — K Means

Moving on to the most exciting part — Model Building! We will be using the K Means Clustering Machine learning algorithm to cluster similar neighborhoods together. There are many different cluster sizes that we can select, we will be going with the number of clusters as 5 to keep it as optimized as possible.

# set number of clusters
k_num_clusters = 5

London_grouped_clustering = London_grouped.drop('Neighbourhood', 1)

# run k-means clustering
kmeans_london = KMeans(n_clusters=k_num_clusters, random_state=0).fit(London_grouped_clustering)

Our model has labeled each of the neighborhoods, we add the label into our dataset.

neighborhoods_venues_sorted_london.insert(0, 'Cluster Labels', kmeans_london.labels_ +1)

We then join London_merged dataframe with our neighborhood venues sorted dataframe to add latitude & longitude for each of the neighborhoods to prepare it for visualization.

london_data = london_merged

london_data = london_data.join(neighborhoods_venues_sorted_london.set_index('Neighbourhood'), on='borough')


Labeled Clustered dataset of London

Visualizing the clustered Neighborhoods

Our data is processed, missing data is collected and compiled. The Model is built. All that’s remaining is to see the clustered neighborhoods on the map. Again, we use Folium package to do so.

We drop all the NaN( Not a Number) values to prevent data skew

london_data_nonan = london_data.dropna(subset=['Cluster Labels'])

Map of clustered neighborhoods of London

Map of clustered neighborhoods of Paris

Examining our Clusters

We could examine our clusters by expanding on our code using the Cluster Labels column:

Cluster 1

london_data_nonan.loc[london_data_nonan['Cluster Labels'] == 1, london_data_nonan.columns[[1] + list(range(5, london_data_nonan.shape[1]))]]

Cluster 2

london_data_nonan.loc[london_data_nonan['Cluster Labels'] == 2, london_data_nonan.columns[[1] + list(range(5, london_data_nonan.shape[1]))]]

Cluster 3

london_data_nonan.loc[london_data_nonan['Cluster Labels'] == 3, london_data_nonan.columns[[1] + list(range(5, london_data_nonan.shape[1]))]]

Cluster 4

london_data_nonan.loc[london_data_nonan['Cluster Labels'] == 4, london_data_nonan.columns[[1] + list(range(5, london_data_nonan.shape[1]))]]

Cluster 5

london_data_nonan.loc[london_data_nonan['Cluster Labels'] == 5, london_data_nonan.columns[[1] + list(range(5, london_data_nonan.shape[1]))]]

Results and Discussion

The neighborhoods of London are very multicultural. There are a lot of different cuisines including Indian, Italian, Turkish, and Chinese. London seems to take a step further in this direction by having a lot of restaurants, bars, juice bars, coffee shops, Fish and Chips shops,s and Breakfast spots. It has a lot of shopping options too with that of the Flea markets, flower shops, fish markets, Fishing stores, clothing stores. The main modes of transport seem to be Buses and trains. For leisure, the neighborhoods are set up to have lots of parks, golf courses, zoos, gyms, and Historic Sites. Overall, the city of London offers a multicultural, diverse, and certainly entertaining experience.

Paris is relatively small in size geographically. It has a wide variety of cuisines and eateries including French, Thai, Cambodian, Asian, Chinese, etc. There are a lot of hangout spots including many Restaurants and Bars. Paris has a lot of Bistros. Different means of public transport in Paris include buses, bikes, boats or ferries. For leisure and sightseeing, there are a lot of Plazas, Trails, Parks, Historic sites, clothing shops, Art galleries, and Museums. Overall, Paris seems like a relaxing vacation spot with a mix of lakes, historic spots, and a wide variety of cuisines to try out.


The purpose of this project was to explore the cities of London and Paris and see how attractive it is to potential tourists and migrants. We explored both the cities based on their postal codes and then extrapolated the common venues present in each of the neighborhoods finally concluding with clustering similar neighborhoods together.

We could see that each of the neighborhoods in both the cities has a wide variety of experiences to offer which is unique in its own way. The cultural diversity is quite evident which also gives the feeling of a sense of inclusion.

Both Paris and London seem to offer a vacation stay or a romantic getaway with a lot of places to explore, beautiful landscapes, amazing food, and a wide variety of cultures. Overall, it’s up-to-the stakeholders to decide which experience they would prefer more and which would more to their liking.

The detailed code is available on GitHub. Thanks for reading!


  1. The Battle of Neighbourhood — My London’s Perspective by Dayo John
  2. The Battle of neighborhoods! What is the best place where can I start my restaurant business in Paris? by Zakaria BOUZIANE
  3. Foursquare API
  4. ArcGIS API