To use the gtrends package for Python3, you'll need to request an API Key from Google. You can do that easily here.
The main function of the package is theo_timeline, a function that pulls and merges data from the Google Trends API and saves a single CSV file for each term (dates on the index and geographies as the columns). Note that multiple files will be created if multiple timeframes are requested. An example script would be:
from gtrendspy import timeline
timeline.theo_timeline(
terms = ['cat', 'cat + food'],
names = ['cat', 'food'],
start = '2018-01-01',
end = '2020-02-01',
timeframe_list = ['month'],
geo_country_list = ['US', 'CA'],
worldwide = False,
timestep_years = 1,
batch_size = 2,
us_states = False,
outpath = "/path/to/output/directory",
creds = "/path/to/info.txt"
)
This script would pull monthly data for the search terms cat and cat + food for both the US and Canada between 1 Jan 2018 and 1 Feb 2020, and then reformat that data into two CSV files (cat_month.csv and food_month.csv, respectively). Each of these CSV files would be formatted such that the first column (timestamp) gives the month and the second and third columns (US and CA) give search queries for the US and Canada, respectively.
This package is intended to be used in conjunction with the gtrendR package for R.
Note: This package assumes use of Python 3.X and pip3. It works on Linux and Windows OS.
(1) pip install from Github
pip install gtrendspy(2) pip install requirements.txt
pip install -r requirements.txt(3) Create a creds file setting your own parameters (DISCOVERY_URL, API_KEY)
Your creds file should look something like this. You'll have access to both once Google approves your request for an API Key.
DISCOVERY_URL = "XXX"
API_KEY = "XXX"
(4) Begin using the package! The merged output will be in the directory you named as outpath.
For now, this code has no license. That means that Theodore L Caputi (Copyright 2020) retains all rights over all code in this package, and unauthorized use, including reuse, modification, and distribution, without written permission of the copyright holder for each specific use is forbidden. This will likely change soon.