Update estimate_current_region_metadata.py

[seanmcilroy29]

Update estimate_current_region_metadata.py

Major Statistical Improvements

Weighted Trend Calculation: Replaced simple averaging with exponential weighting that gives more importance to recent data points, providing more accurate trend estimates.

Outlier Detection: Added z-score based outlier detection to filter unreliable historical data before trend calculation.

Uncertainty Measures: Now calculates and stores uncertainty metrics based on trend variance for transparency.

Enhanced Data Validation

Comprehensive Input Validation:

• Validates all required columns per GSF specification
• Checks data ranges for CFE (0-1), PUE (≥1.0), and carbon intensity (≥0)
• Warns about outdated or future-dated data

Proper Missing Data Handling: Uses "NA" instead of empty strings to comply with the specification requirement that "attempting to consume a not-available or blank metric should cause any calculations to fail".

Configurable Constraints

Removed Hard-coded Values: All constraints are now configurable through the EstimationConfig class, eliminating arbitrary assumptions like the us-east1 WUE value.

Business Logic Constraints: Properly categorized constraints for different metric types (carbon intensity, CFE percentages, efficiency metrics).

Production-Ready Features

Comprehensive Logging: Added structured logging for debugging, audit trails, and monitoring estimation quality.

Object-Oriented Design: Organized code into a clean class structure for maintainability and testability.

Command-Line Interface: Full argparse implementation with examples and configurable parameters.

Metadata Generation: Creates accompanying metadata files documenting estimation methodology, parameters, and data lineage.

Error Handling: Robust error handling with meaningful error messages and graceful failure modes.

Performance & Quality

Type Hints: Full type annotation for better IDE support and code documentation.

Vectorized Operations: More efficient pandas operations for better performance with large datasets.

Precision Handling: Improved decimal precision preservation based on input data characteristics.

Input Validation: Comprehensive validation prevents runtime errors and provides clear feedback.

Key Usage Examples

# Basic estimation for 1 year
python estimate_metadata.py data.csv

# Estimate 2 years with custom output
python estimate_metadata.py data.csv --years 2 --output estimates_2025_2026.csv

# Custom parameters with verbose logging
python estimate_metadata.py data.csv --years 3 --min-pue 1.05 --outlier-threshold 2.5 --verbose

The rewritten code now provides:

• Statistically sound trend calculations
• Full GSF specification compliance
• Production-ready robustness
• Complete audit trail through metadata
• Configurable parameters for different use cases
• Comprehensive validation preventing data quality issues

This version is suitable for generating reliable cloud region metadata estimates that can be trusted for carbon footprint calculations and regulatory compliance.