Link to Live Tabluea Dashboard: https://tabsoft.co/3axmz8j
Version two of the Google Search Trends API.
After completing the initial testing of the first version, the next step was to create a series of functions in line with the principles of object-oriented programming.
Initially was created in a linear format as a proof of concept. Later split the code into five main functions.
- create_unique_directory():
The purpose of this function is to create a unique directory based on the current date and time stamp. Executes each time the process is run and used to store the payloads received by the following three functions.
- get_interest_over_time(month):
The purpose of this function is to retrieve the interest over time, a time-series data for the main keyword specified in the keyword collection. Either with a time frame of 1 or 3 months. The time frame is specified though the month parameter. The API returns the data in the format of a DataFrame.
- get_related_queries(month):
The purpose of this function is to retrieve search terms related to the main keyword specified in the keyword collection. Either with a time frame of 1 or 3 months. The time frame is specified though the month parameter. The API returns the data in the format of a data dictionary than is converted to a DataFrame. The dictionary comes with top keywords and rising keywords. The final DataFrame consists of top keywords portion of the dictionary.
- get_interest_by_region(month):
The purpose of the function is to retrieve interest by region within the specified geography for the main keyword specified in the keyword collection. Either with a time frame of 1 or 3 months. The time frame is specified though the month parameter. The API returns the data in the format of a DataFrame.
- concat_payloads(pd_payload_one, pd_payload_two):
The purpose of this function is the concatenate two DataFrames into one single DataFrame then stored in a CSV file.
The keyword collection is specified as follows. The keyword list collection can take up to 5 keywords. kw_list = ["covid"]
The time frame is determined by the month parameters as per below, returns the following output in the format required by the Google API. if month == 1: time_frame = "today 1-m" elif month == 3: time_frame = "today 3-m"
Besides specifying the time frame, we also need to specify the geography through the geo parameter. pytrends.build_payload(kw_list, cat=0, timeframe=time_frame, geo='GB', gprop='')
For this particular project, each of the get_ functions is executed twice. Once to get 30-days of data and a second time to get 90-days of data. Both DataFrames feed into to the concat_payloads() with the final output stored in a CSV file.
The code itself, well commented with each step of the process printed out on screen for verification along with other specific actions taken such as creating the directory and saving the payloads to CSV file.