Google Trends

Google Trends#

import os
os.chdir("../../")

import inspect
import numpy as np
import pandas as pd
from src.google_trends import GT


from bokeh.layouts import column
from bokeh.models import ColumnDataSource, CustomJS, Select, HoverTool
from bokeh.plotting import figure, show, output_notebook

output_notebook()

Loading BokehJS ...

The following code was modified from Development Data Partership. See example code.

print(inspect.getsource(GT))

class GT:
    def __init__(self, _GOOGLE_API_KEY):
        self.service = build(
            serviceName=SERVICE_NAME,
            version=SERVICE_VERSION,
            discoveryServiceUrl=_DISCOVERY_SERVICE_URL,
            developerKey=_GOOGLE_API_KEY,
            cache_discovery=False)
        self.block_until = None

    def get_health_trends(self, terms, timelineResolution="month"):
        graph = self.service.getTimelinesForHealth(
            terms=terms,
            timelineResolution=timelineResolution
        )

        try:
            response = graph.execute()
            return response

        except HttpError as http_error:
            data = json.loads(http_error.content.decode('utf-8'))
            code = data['error']['code']
            reason = data['error']['errors'][0]['reason']
            if code == 403 and reason == 'dailyLimitExceeded':
                self.block_until = datetime.combine(
                    date.today() + timedelta(days=1), dtime.min)
                raise RuntimeError('%s: blocked until %s' %
                                   (reason, self.block_until))
            logging.warning(http_error)
            return []

    def get_graph(self, terms,
                  restrictions_geo,
                  restrictions_startDate="2004-01"):
        graph = self.service.getGraph(
            terms=terms,
            restrictions_geo=restrictions_geo,
            restrictions_startDate=restrictions_startDate
        )

        try:
            response = graph.execute()
            return response

        except HttpError as http_error:
            logging.warning(http_error)
            return []

    def get_top_topics(self, term,
                       restrictions_geo,
                       restrictions_startDate="2004-01"):
        graph = self.service.getTopTopics(
            term=term,
            restrictions_geo=restrictions_geo,
            restrictions_startDate=restrictions_startDate
        )
        try:
            response = graph.execute()
            return response
        except Exception as e:
            logging.warning(e)
            return []

    @staticmethod
    def to_df(result: json) -> pd.DataFrame:
        df = pd.json_normalize(result["lines"], meta=[
                               "term"], record_path=["points"])
        if "date" in df.columns:
            df["date"] = pd.to_datetime(df["date"])

        return df

folder = os.getcwd() + "/data/tourism/trends/"
filepaths = [folder + path for path in os.listdir(folder)]
GoogleAPI = os.getenv("GoogleAPIkey")

trends = pd.DataFrame()
for file in filepaths:
    if "readme" not in file:
        country_trend = (pd.read_csv(file).drop("Unnamed: 0", axis=1)
                           .iloc[:, [0, -3, -2, -1]])
        country_trend.columns = [col.replace(" ", "_").lower()
                                 for col in country_trend.columns]
        country_trend["date"] = pd.to_datetime(country_trend["date"])

        if trends.empty:
            trends = country_trend
        else:
            trends = trends.merge(country_trend, how="left", on="date")

display(trends.head(5))

	date	palau_flights	palau_hotel	palau_travel	solomon_islands_flights	solomon_islands_hotel	tonga_flights	tonga_hotel	tonga_travel	vanuatu_flights	vanuatu_hotel	vanuatu_travel	fiji_flights	fiji_hotel	fiji_travel	samoa_flights	samoa_hotel	samoa_travel
0	2004-01-01	0.000000	4.374890	1.137126	0.000000	1.511560	2.880180	1.414027	1.422723	0.000000	1.641851	4.364237	7.736326	12.222561	21.884040	0.000000	1.130191	1.297947
1	2004-02-01	0.000000	4.834610	0.000000	0.923086	0.883077	4.318772	6.542797	0.000000	0.896698	1.434310	2.362955	8.971634	13.282344	19.926081	4.138847	5.995897	1.808004
2	2004-03-01	0.000000	6.227588	1.372316	0.000000	1.216019	0.000000	4.433156	0.000000	4.973685	2.122763	4.711485	11.508566	12.604433	18.692017	0.000000	0.000000	2.331247
3	2004-04-01	0.000000	4.139832	0.000000	0.994321	2.589010	0.000000	0.927113	0.000000	1.273081	0.000000	3.575172	6.667417	8.880059	15.798739	0.000000	2.146069	0.000000
4	2004-05-01	0.900479	6.006549	0.000000	0.000000	0.000000	0.742281	1.906869	3.958648	0.980399	1.270780	2.740211	9.652304	13.673188	15.091649	0.000000	2.409763	0.000000