Skip to content

Changelog

v0.1.3 – 2025-09-20

New Features:

  • MovingWindowSmoothing class for applying moving window smoothing to time series data.
  • Helps in reducing noise and filling short cloud-induced gaps.
  • TemporalInterpolation class for interpolating gaps in satellite time series.
  • RegularTimeseries class for converting irregular satellite time series into regular intervals.
  • Flexible frequency specification (e.g., 5-day, 10-day, 15-day).

Improvements:

  • Expanded documentation with notebook examples for smoothing, interpolation, and regularization workflows.

v0.1.2 – 2025-09-01

New Features:

  • TimeseriesExtractor class for downloading per-feature CSV time series from Earth Engine ee.ImageCollection.
  • Point geometries: uses getRegion, no reducer required.
  • Polygon/MultiPolygon geometries: supports reducers for aggregation.
  • Exports directly to CSV for downstream analysis.
  • HarmonicRegression class for performing harmonic regression on Earth Engine ee.ImageCollection.
  • Computes harmonic coefficients with configurable order and base frequency.
  • Provides phase and amplitude outputs for phenology and seasonal dynamics analysis.
  • Supports fitted harmonic time series generation for visualization or modeling.

Improvements:

  • Updated docstrings for clarity and alignment with function behavior.
  • Documentation expanded to include harmonic regression and time series extraction workflows.

v0.1.1 – 2025-08-06

New Features:

  • ImagePatchExtractor class added for efficient extraction of image patches from Earth Engine ee.Image objects using local sample points as GeoDataFrame.
  • Supports multiple export formats: png, jpg, GEO_TIFF, and others.
  • Fully parallelized with configurable number of processes.
  • Uses a specified identifier column for naming output files.
  • Automatically handles patch sizing via dimensions and buffer parameters.

Improvements - Improved documentation and example notebooks:

v0.1.0 - 2025-07-29

Improvements:

  • Improved initial project scaffolding and modular structure.
  • Enhanced configuration for easier customization and extension.

New Features:

  • Added preprocessing module with various image scaling options:
  • MeanCentering
  • MinMaxScaler
  • StandardScaler
  • RobustScaler
  • Added analysis module including easy implementation of PCA with explained variance calculation.
  • Added new example notebooks.