I refer to the index of googleplaces : https://googlemaps.github.io/google-maps-services-python/docs/index.html
You could also find a lot of other useful function there.
Install the google maps
pip install -U googlemaps
Initiate Google Map Client
# make sure in your env, there are necessary libs.
import googlemaps
import time
import pandas as pd
google_maps_key = '{your token here}'
gmaps = googlemaps.Client(google_maps_key)
Get data from Google Map
What I want to tell readers and it really makes me confused for quite a long time is : Google Map can provide **at most 60 results** once!!!!
No way can change this rule except you are willing to pay for its service.
What’s more, pls remember to add time.sleep(5) between every next page request(because every page there are only at most 20 results and you can only request next page at most 3 times. That is why we can only get at most 60 results)
###########################gmaps.places####################################
# _add_result_to_df is used to add the most 60 results to the dataframe
# @ data_df,there is a limitation in Google_map, so it is really a pity to get all the data from
# @ data_df: the dataframe used to save detailed information
# @ params:
def _add_result_to_df(data_df,params):
while True:
x = gmaps.places(**params)
# print (x['results'][0]['name'])
print (len(x['results']))
time.sleep(2)
if 'next_page_token' in x:
params['page_token'] = x['next_page_token']
for item in x['results']:
data_df = data_df.append({'name': item['name'], 'lat': item['geometry']['location']['lat'],
'lng': item['geometry']['location']['lng'],'type':params['query']}, ignore_index=True)
else:
for item in x['results']:
data_df = data_df.append({'name': item['name'], 'lat': item['geometry']['location']['lat'],
'lng': item['geometry']['location']['lng'],'type':params['query']}, ignore_index=True)
break
return data_df
# query_google_map is used to search for a concrete text like "school in Copenhagen"
# @ query_text: the content of what you want to search
# TIP: the maxium length is 60 limited by GoogleMap
def query_google_map(query_text):
data = pd.DataFrame(columns=['name','lat','lng','type'])
params = {
'query': query_text
}
data = _add_result_to_df(data,params)
return data
# ######################gmaps.places_autocomplete####################################
def _add_detail_and_sum_result(data_detail,data_sum,params):
count = 0
while True:
x = gmaps.places_autocomplete(**params)
print(x)
count += len(x)
# print (x['results'][0]['name'])
time.sleep(2)
if 'next_page_token' in x:
params['page_token'] = x['next_page_token']
for item in x:
data_detail = data_detail.append(
{'name':item['description'],
'type':params['input_text'],
'lat_based':params['location']['lat'],
'lng_based':params['location']['lng']}, ignore_index=True)
else:
for item in x:
data_detail = data_detail.append(
{'name':item['description'],
'type':params['input_text'],
'lat_based':params['location']['lat'],
'lng_based':params['location']['lng']}, ignore_index=True)
break
data_sum = data_sum.append(
{
'lat_based':params['location']['lat'],
'lng_based':params['location']['lng'],
'type':params['input_text'],
'sum':count
},ignore_index=True
)
return data_detail,data_sum
# query_with_lat_lng_radius_type is used
def places_autocomplete_with_lat_lng_radius_type(lat,lng,radius,query_text):
data_detail = pd.DataFrame(columns=['name','type','lat_based','lng_based'])
data_sum = pd.DataFrame(columns=['lat_based','lng_based','type','sum'])
params = {
'input_text': query_text,
'location':{"lat": lat, "lng": lng},
'radius': radius,
'strict_bounds':True
}
data_detail,data_sum = _add_detail_and_sum_result(data_detail,data_sum,params)
return data_detail,data_sum
# ############################gmaps.places####################################
# the type of google_map:https://developers.google.com/maps/documentation/places/web-service/supported_types
def _add_result_to_df_unique(data_df,params):
while True:
x = gmaps.places(**params)
time.sleep(2)
if 'next_page_token' in x:
params['page_token'] = x['next_page_token']
for item in x['results']:
temp_list = {
'name': item['name'],
'lat': item['geometry']['location']['lat'],
'lng': item['geometry']['location']['lng'],
'type':params['query']
}
if not ((data_df['name'] == temp_list['name'])&(data_df['lat'] == temp_list['lat'])&(data_df['lng'] == temp_list['lng'])).any():
data_df = data_df.append(temp_list, ignore_index=True)
else:
for item in x['results']:
temp_list = {
'name': item['name'],
'lat': item['geometry']['location']['lat'],
'lng': item['geometry']['location']['lng'],
'type':params['query']
}
if not ((data_df['name'] == temp_list['name'])&(data_df['lat'] == temp_list['lat'])&(data_df['lng'] == temp_list['lng'])).any():
data_df = data_df.append(temp_list, ignore_index=True)
break
return data_df
# data = pd.DataFrame(columns=['name','lat','lng','type'])
def query_place_with_lat_lng_type(lat,lng,result_data):
params = {
'query': address_type,
'location':{"lat": lat, "lng": lng}
}
result_data = _add_result_to_df_unique(result_data,params)
return result_data
# ########################gmaps.places_nearby############################
def _add_nearby_result_to_df(result_data,params):
while True:
x = gmaps.places_nearby(**params)
time.sleep(2)
if 'next_page_token' in x:
params['page_token'] = x['next_page_token']
for item in x['results']:
temp_list = {
'name': item['name'],
'lat': item['geometry']['location']['lat'],
'lng': item['geometry']['location']['lng'],
'type':item['types']
}
if not ((result_data['name'] == temp_list['name'])&(result_data['lat'] == temp_list['lat'])&(result_data['lng'] == temp_list['lng'])).any():
result_data = result_data.append(temp_list, ignore_index=True)
else:
for item in x['results']:
temp_list = {
'name': item['name'],
'lat': item['geometry']['location']['lat'],
'lng': item['geometry']['location']['lng'],
'type':item['types']
}
if not ((result_data['name'] == temp_list['name'])&(result_data['lat'] == temp_list['lat'])&(result_data['lng'] == temp_list['lng'])).any():
result_data = result_data.append(temp_list, ignore_index=True)
break
return result_data
def nearby_with_lat_lng_radius(lat,lng,radius,result_data):
params = {
'radius': radius,
'location':{"lat": lat, "lng": lng}
}
result_data = _add_nearby_result_to_df(result_data,params)
return result_data
Send Request and write to Csv
df_search_result = pd.DataFrame(columns=['name','lat','lng','type'])
# df_search_result = nearby_with_lat_lng_radius(55.708,12.569,300,df_search_result)
# df_search_result = nearby_with_lat_lng_radius(55.657,12.568,300,df_search_result)
for i in range(101,1000):
# print(df_geo.loc[i,:][2],df_geo.loc[i,:][1])
print(i)
df_search_result = nearby_with_lat_lng_radius(df_geo.loc[i,:][2],df_geo.loc[i,:][1],300,df_search_result)
if (i%100==0 and i != 0):
time.sleep(5)
print("len:",len(df_search_result))
df_search_result.to_csv("search_result_101_1000.csv",index=False,mode='w+')
print('write success!')
The example I got is :
