Metadata-Version: 2.4
Name: FastArray
Version: 0.1.1
Summary: A lightweight Python package for preprocessing geospatial optical imagery for machine learning
Author-email: "Isah Abdul-Azeez." <isahabdul16@gmail.com>
License: MIT
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy
Requires-Dist: xarray
Requires-Dist: aiofiles
Requires-Dist: matplotlib
Requires-Dist: torch
Requires-Dist: rioxarray
Requires-Dist: pyproj
Requires-Dist: shapely
Dynamic: license-file

# fastarray

**fastarray** is a fast, lightweight Python package for preprocessing geospatial optical imagery (e.g., Sentinel-2) to prepare it for machine learning workflows. It simplifies operations like index computation (NDVI, NDWI, etc.), reprojection, patching, and transformation into PyTorch-ready tensors—while preserving geospatial coordinates.

---

## 🚀 Features

- ✅ Compute vegetation and water indices (NDVI, NDWI, etc.)
- ✅ Read raster data into coordinate-aware arrays
- ✅ Clip and reproject rasters using geometries (e.g., GeoJSON or Shapely)
- ✅ Split imagery into fixed-size patches
- ✅ Convert to PyTorch tensors for ML pipelines
- ✅ Fully Pythonic (no heavy GDAL/C++ dependencies)
- ✅ Fast and async-friendly where possible

---

## 📦 Installation

```bash
pip install fastarray

