English version | Chinese version README.md

Documents | Pypi

Overview

• Overview
• Introduction
• How to Stream GA4 Data to Bigquery in Real-time?
• Features
• How to Use?
  • Download the Package
  • Connect to Your Bigquery
  • Connect to GA4 Tables
  • Evaluate Query Cost with dry run
  • Analyze User Properties
  • Analyze Device Properties
  • Analyze Events

---

Introduction

• pyGA4 is a Python toolkit designed for extracting, processing, and analyzing data from Google Analytics 4 (GA4).
• Whether you're a digital marketing professional, a data analyst, or anyone interested in gaining insights from GA4 data, this package simplifies the process of working with your GA4 data.

How to Stream GA4 Data to Bigquery in Real-time?

First, we assume that everyone has already integrated GA4 data into their respective platform websites (there are many online tutorials).

Next, we will use a free third-party service to stream data into Bigquery. For detailed instructions, please refer to the official documentation.

If successful, you will see tables in Bigquery similar to the following (analytics_xxxx), ref:

Features

• Query Cost Estimation: Provides the Bigquery dry run feature to estimate query cost before execution.
• Data Extraction: Easily connect to your GA4 property, retrieve data, and save it for analysis.
• Data Preprocessing: Prepare and clean your GA4 data for analysis with built-in data preprocessing functions.
• Custom Queries: Execute custom queries to filter and aggregate data based on your specific needs.
• Data Analysis: Perform various types of analysis, including user behavior analysis, conversion tracking, and more.
• Data Visualization: Create informative visualizations and reports to communicate your findings effectively.
• Simple Integration: Seamlessly integrate pyGA4 into your data pipeline or analytics workflow.

How to Use?

For more features, please refer to the package documentation.

Download the Package

pip install pyga4

Connect to Your Bigquery

from google.cloud import bigquery

client = bigquery.Client()
# Or you can use:
# client = bigquery.Client.from_service_account_json(
#    './private/service-project-data-dev-01d11c742ba1.json'
# )

Connect to GA4 Tables

from pyga4.model import Ga4Table

# Use your project_id, dataset_name (analytics_xxxx)
ga4_table = Ga4Table(client, PROJECT_ID, DATASET_NAME)

# Show the tables list in dataset, e.g., analytics_date1, analytics_date2
table_id_list = ga4_table all_tables_list
print(table_id_list)

# Select the table you want to analyze
ga4_table.table_id = 'events_intraday_20200812'

Evaluate Query Cost with dry run

    # Query with dry run:
    ga4_table.query_config.dry_run = True
    query = f"""
    SELECT event_timestamp FROM `<project_id>.<dataset_name>.<data_table>`
    """
    results = ga4_table.query(query) # return None, but you can see the query usage!

Analyze User Properties

Query User ID and Country List

# User attribute
user_id_list = ga4_table.user_id_list
user_country_list = ga4_table.geo_country_list

Query User ID and Country Distribution

from pyga4.analytic import UserAnalytic

# UserAnalytic
user_analytic = UserAnalytic(ga4_table)
countries_dist = user_analytic.countries_distribution
userid_dist = user_analytic.user_id_distribution

Analyze Device Properties

# DeviceAnalytic
device_analytic = DeviceAnalytic(ga4_table)
mobile_brand_dist = device_analytic.mobile_brand_distribution

Analyze Events

# EventAnalytic
event_analytic = EventAnalytic(ga4_table)
page_loc_dist = event_analytic.pages_distribution

Back to top