Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/terrafloww/rasteret/llms.txt

Use this file to discover all available pages before exploring further.

Beat Cold Starts with Rasteret

Index-first access to cloud-native GeoTIFF collections. Parse headers once, cache in Parquet, read pixels 20x faster for ML and geospatial analysis.

Get StartedAPI Reference

Why Rasteret?

20x Faster

No cold-start penalty. Header metadata lives in Parquet, not behind HTTP. Same speed every time.

Zero Downloads

Work with terabytes of cloud imagery while storing only megabytes of metadata locally.

No STAC at Training Time

Query STAC once at setup. Zero API calls during training loops, no rate-limiting risk.

Quick Start

1

Install Rasteret

Install via pip or uv. Requires Python 3.12+.
uv pip install rasteret
2

Build a Collection

Pick a dataset from the built-in catalog and create a collection with your bbox and date range.
import rasteret

collection = rasteret.build(
    "earthsearch/sentinel-2-l2a",
    name="s2_training",
    bbox=(77.5, 12.9, 77.7, 13.1),
    date_range=("2024-01-01", "2024-06-30"),
)
Headers are parsed once and cached as Parquet. Subsequent runs load in milliseconds.
3

Use Your Data

Get data as NumPy arrays, xarray Datasets, or TorchGeo-compatible datasets for ML training.
# For ML training
dataset = collection.to_torchgeo_dataset(
    bands=["B04", "B03", "B02", "B08"],
    chip_size=256,
)

# For analysis
ds = collection.get_xarray(
    geometries=(77.55, 13.01, 77.58, 13.08),
    bands=["B04", "B08"],
)
ndvi = (ds.B08 - ds.B04) / (ds.B08 + ds.B04)

Explore by Topic

Core Concepts

Understand index-first retrieval, collections, and the dataset catalog

Building Collections

Learn how to build collections from STAC APIs, GeoParquet, or custom sources

ML Training

Integrate with TorchGeo for satellite imagery ML pipelines

Data Analysis

Query and analyze geospatial data with xarray and NumPy

Cloud Authentication

Configure backends for Planetary Computer, AWS requester-pays, and Earthdata

Dataset Catalog

Browse 12+ built-in datasets including Sentinel-2, Landsat, NAIP, and more

Built-in Datasets

Rasteret ships with a catalog of 12+ datasets. Each entry includes license metadata and a commercial_use flag.
rasteret datasets list
  • Sentinel-2 Level-2A (Element84, Planetary Computer)
  • Landsat Collection 2 Level-2 (Element84)
  • NAIP (Element84)
  • Copernicus DEM (30m, 90m)
  • ESRI 10m Land Use/Land Cover (Planetary Computer)
  • ALOS World 3D 30m DEM (Planetary Computer)
  • NASADEM (Planetary Computer)
  • ESA WorldCover (Planetary Computer)
  • USDA Cropland Data Layer (Planetary Computer)
  • AlphaEarth Foundation Embeddings (Annual, global)
See the Dataset Catalog for full details.
The catalog is community-driven. Each entry is ~20 lines of Python. One PR adds a dataset for all users.

Integrations

TorchGeo

Drop-in GeoDataset for ML training with TorchGeo samplers and DataLoaders

xarray

Query and analyze multi-dimensional geospatial data with xarray

STAC

Build collections from any STAC API endpoint

Community & Support

GitHub Discussions

Ask questions, share workflows, and discuss features

Discord Community

Join the community for real-time chat and support

Contributing

Add datasets, improve docs, or contribute to the codebase

Issues & Bug Reports

Report bugs and request features on GitHub

Ready to accelerate your geospatial workflows?

Start building with Rasteret today and say goodbye to cold starts.

Get Started →