SeasonalImpute

A Python package for imputing missing values in time series data using a seasonal weighted average approach.

Installation

pip install SeasonalImpute

Usage

import numpy as np
from SeasonalImpute import SeasonalWeightedAverageImputation

# Example data
data = np.array([1.0, np.nan, 3.0, 1.0, np.nan, 3.0])

# Impute with seasonality
imputer = SeasonalWeightedAverageImputation(window=3, seasonality={2: 0.5})
imputed_data = imputer(data)
print(imputed_data)

Features

• Imputes missing values using nearby values and seasonal patterns.
• Customizable window size and seasonal weights.
• Built on gluonts and numpy for robust time series handling.

Development

To contribute:

1. Clone the repository:

   git clone https://github.com/hanifkia/SeasonalImpute.git

2. Install dependencies:

   pip install -e .[dev]

3. Run tests:

   pytest

License

MIT License. See LICENSE for details.