tssynth 0.1.2

A library for generating synthetic time series data

tssynth

Generate realistic synthetic time series data with customizable patterns, seasonality, and noise components. Perfect for testing time series algorithms, machine learning models, or generating sample datasets.

Features

✨ Flexible Generation: Create time series with various components:

• Linear, exponential, or custom trends
• Multiple seasonality patterns
• Configurable noise distributions
• Anomaly injection

🚀 Easy to Use: Simple, intuitive API built on numpy and pandas

import tssynth as ts

# Generate a time series with multiple components
data = ts.generate(
    length=365,                    # One year of daily data
    trend="exponential",           # Exponential growth
    seasonality=["weekly", "yearly"],
    noise_type="gaussian",
    noise_level=0.05
)

Installation

pip install tssynth

Documentation

For detailed usage examples and API reference, visit our documentation.

Quick Examples

# Generate a simple trend
basic_trend = ts.generate(length=100, trend="linear")

# Add seasonality and anomalies
complex_ts = ts.generate(
    length=1000,
    trend="exponential",
    seasonality="monthly",
    anomalies={"frequency": 0.01, "magnitude": 3.0}
)

# Export to pandas DataFrame
df = complex_ts.to_dataframe()

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Format code
black .
isort .

License

MIT License - See LICENSE file for details.

Author

Will Judge ([e-mail])