clustering.py

# -*- coding: utf-8 -*-
"""Clustering.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1oHWnSAo_tOYKekEWU-7QTdEgmkI4to2L
"""

!pip install googlemaps

import pandas as pd
import requests
import numpy as np
import urllib.parse
import re
import json
import googlemaps
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans

attractions = ['Eiffel Tower','Louvre Museum','Notre Dame Cathedral','Basilica Sacre Coeur','Arc de Triomphe','Disneyland Paris','La Vallee Village','Pantheon','Musee d\'Orsay']
gmaps = googlemaps.Client(key='AIzaSyCab__7ZxRiq-b5OXDkRB-3Z9zpSn_SmsE')

myDict = {}
myDict['location'] = attractions
myDict['lat'] = []
myDict['lon'] = []
arr = []

for add in attractions:
    address = add + ' France'
    mapurl = 'https://nominatim.openstreetmap.org/search/' + urllib.parse.quote(address) +'?format=json'

    response = requests.get(mapurl).json()
    try:
        lat = response[0]["lat"]
        lon = response[0]["lon"]
    except:
        lat= None
        lon= None
    
    myDict['lat'].append(lat)
    myDict['lon'].append(lon)
    arr.append([lat,lon])

df = pd.DataFrame(myDict)

df['lat'] = df['lat'].astype(float)
df['lon'] = df['lon'].astype(float)

kmeans = KMeans(n_clusters=3, random_state=0).fit(arr)

df['group'] = kmeans.labels_

df.sort_values('group')

def getShortest(start,visited,grp):
    temp_lst = [loc for loc in grp['location'] if loc not in visited]
    dist_lst = []
    for loc in temp_lst:
        dist = gmaps.distance_matrix(start,loc)['rows'][0]['elements'][0]
        distance = dist['distance']['text']
        distance = distance.replace(' km','')
        distance = distance.replace(',','')
        distance = float(distance)
        dist_lst.append((loc,distance))
    return sorted(dist_lst, key=lambda x: x[1])[0]

def getPath(i):
    grp = df[df['group']==i]
    grp = grp.reset_index(drop = True)
    if len(grp)>1:
        lst = []
        start = grp['location'][0]
        lst.append(start)
        cur = start
        tot_dist = 0
        for i in range(len(grp)-1):
            next,dist_ = getShortest(cur,lst,grp)
            lst.append(next)
            tot_dist +=dist_
            cur = next
        return lst,tot_dist
    else:
        return [grp['location'][0]],0

for i in range(3):
  print(getPath(i))