delta-lens 0.1.6

A tool for comparing large datasets using DuckDB

 ____       _ _        _                    
|  _ \  ___| | |_ __ _| |    ___ _ __  ___ 
| | | |/ _ \ | __/ _` | |   / _ \ '_ \/ __|
| |_| |  __/ | || (_| | |__|  __/ | | \__ \
|____/ \___|_|\__\__,_|_____\___|_| |_|___/
                                        

DeltaLens - Data Comparison Tool

DeltaLens is a powerful tool for comparing large datasets using the power of DuckDB as the comparison engine. It supports data transformations, automated field-level matching, and detailed comparison reporting.

flowchart LR
    Trades_1@{ shape: doc, label: "new_system_trades.csv" }
    Trades_2@{ shape: doc, label: "lagecy_system_trades.csv" }
    Config@{ shape: doc, label: "config.json" }
    Config-->TableQueryGenerator
    Trades_2 -->|load|DuckDB
    Trades_1 -->|load|DuckDB
    subgraph DeltaLens.py
        TableQueryGenerator@{ shape: subproc, label: "QueryGenerator" }
        TableQueryGenerator-->|generate compare queries|DuckDB
        DuckDB@{ shape: lin-cyl, label: "DuckDB" }
        DuckDB-->Exporter
        Exporter@{ shape: subproc, label: "Exporter" }
       
    end
    Exporter-->|export|Sqlite
    Sqlite@{ shape: lin-cyl, label: "results.sqlite" }

Features

• Compare CSV datasets with configurable primary keys
• Apply SQL transformations to data before comparison
• Generate detailed field-level match statistics
• Export results to SQLite and CSV for analysis
• Support for larger than memory datasets
• Support for reference datasets
• Docker support for containerized execution
• CLI and Python API interfaces
• Data Pipeline and CI/CD friendly

Installation

Install from PyPI:

pip install delta-lens

see data_compate.ipynb for example.

Basic Usage

Command Line Interface

# Basic comparison
deltalens --config data/compare.config.json --run-name daily_compare

# Full options
deltalens \
  --config data/compare.config.json \
  --run-name daily_compare \
  --output-dir ./results \
  --persistent \
  --continue-on-error \
  --export-sqlite \
  --export-csv \
  --export-sampling-threshold 5000 \
  --export-mismatches-only \
  --log-level DEBUG

Or Pull the Docker image:

docker run unclepaul84/deltalens:latest

Using DeltaLens in Jupyter Notebooks

DeltaLens can be used interactively in Jupyter notebooks for data comparison analysis. See data_compate.ipynb

Configuration

Create a compare.config.json file:

{  
    "defaults":{},
    "entities": [
        {
            "entityName":"trade",
            "leftSide": {
                "title": "legacy",
                "inputFile":"data/legacy_system_trades.csv"
                
                

            },
            "rightSide": {
                "title": "new",
                "inputFile":"data/new_system_trades.csv"
            },
            "primaryKeys": ["trade_id"]
        
        }     
    ]
}

Environment Variables

Variable	Description	Default
DELTALENS_CONFIG	Path to config file	compare.config.json
DELTALENS_RUN_NAME	Name for comparison run	compare_YYYY-MM-DD
DELTALENS_OUTPUT_DIR	Output directory	.
DELTALENS_PERSISTENT	Use persistent storage	false
DELTALENS_EXPORT_SQLITE	Export to SQLite	true
DELTALENS_EXPORT_SAMPLING_THRESHOLD	rowcount at which to start sampling	10000
DELTALENS_EXPORT_CSV	export to gzipped csv	true
DELTALENS_EXPORT_MISMATCHES_ONLY	Export mismatched rows only	true

Output Files

The tool generates several output files:

• [run_name].duckdb: DuckDB database with comparison results (if persistent mode enabled)
• [run_name].sqlite: SQLite export of comparison results (if enabled)

Resulting Tables include:

• entity_compare_results: Overall comparison summary
• [entity]_compare: Detailed record-level comparison
• [entity]_compare_field_summary: Field-level match statistics

Development

Docker

# Run with docker-compose
docker-compose up

# Run with custom arguments
docker-compose run deltalens --run-name custom_run --log-level DEBUG

Generating Sample Data

DeltaLens includes a script to generate sample trade data for testing and demonstration purposes.

Sample Data Generator

The script creates two CSV files with randomized trade data:

• legacy_system_trades.csv: Original trade data with modifications
• new_system_trades.csv: Copy of original data with known differences

cd data
# Generate sample data (creates 2GB files by default)
python create_test_datasets.py

# Install development dependencies
pip install -r requirements.txt
pip install -r requirements-dev.txt

# Run tests
pytest -v

# Run tests with coverage
pytest --cov=delta_lens -v

License

MIT License

Contributing

1. Fork the repository
2. Create a feature branch
3. Submit a pull request