ga4-analytics-processor 0.2.0

A Python library for processing Google Analytics 4 data

GA4 Analytics Processor

A Python library for processing Google Analytics 4 data with built-in data transformations and channel grouping logic.

Installation

pip install ga4-analytics-processor

Quick Start

from ga4_analytics import GA4Processor
from datetime import datetime, timedelta

# Initialize processor
processor = GA4Processor(
    credentials_path="/path/to/your/credentials.json",
    property_id="your_ga4_property_id"
)

# Define date range
end_date = datetime.now().date()
start_date = end_date - timedelta(days=7)

# Fetch and process general report with default dimensions/metrics
df = processor.get_general_report(start_date, end_date)

# Or specify custom dimensions and metrics
custom_dimensions = ["date", "sessionDefaultChannelGroup", "deviceCategory"]
custom_metrics = ["sessions", "totalRevenue", "transactions"]
df_custom = processor.get_general_report(
    start_date, end_date,
    dimensions=custom_dimensions,
    metrics=custom_metrics
)

# For completely custom reports with optional processing
df_raw = processor.get_custom_report(
    start_date, end_date,
    dimensions=custom_dimensions,
    metrics=custom_metrics,
    apply_processing=False  # Get raw data without transformations
)

print(df.head())

Features

• Easy GA4 API integration
• Flexible dimensions and metrics - specify your own or use defaults
• Custom report generation - with optional data processing
• Built-in data transformations and cleaning
• Channel grouping and campaign labeling
• Revenue reallocation logic
• Email campaign data integration
• Configurable data formatting

New in v0.2.0

• ✅ Custom dimensions and metrics in get_general_report()
• ✅ New get_custom_report() method for full flexibility
• ✅ Optional data processing - get raw data or processed data
• ✅ Backward compatible - existing code continues to work

Requirements

• Python 3.8+
• Google Analytics Data API credentials
• pandas, numpy, google-analytics-data

License

MIT License