Skip to content

README.md

Latest commit

 

History

History
79 lines (63 loc) · 3.11 KB

README.md

File metadata and controls

79 lines (63 loc) · 3.11 KB
Raw

taco

Read and write TACO datasets in Python.

pip install taco-eo
python examples/minimal.py

Published wheels include the native TACO reader. Building from the source distribution requires a C++23 compiler, CMake, Ninja, pkg-config, libcurl 7.83 or newer, and OpenSSL 3 or newer.

import taco

samples = taco.read("dataset.zip")
parts = taco.read(["part-0.zip", "part-1.zip"])

dataset = taco.open_dataset("dataset.zip")
targets = dataset.read(files="target.tif")
train = dataset.sql("SELECT * FROM data WHERE \"ml:split\" = 'train'")

Versioned dataset roots select their declared default release without listing remote storage. Use a release URL to open an immutable version directly.

dataset = taco.open_dataset("https://data.source.coop/major-tom/core-dem/")
print(dataset.version)
print(dataset.versions)

previous = taco.open_dataset("https://data.source.coop/major-tom/core-dem/1.0.0/")

export() writes a smaller dataset with the same contract. samples is a PyArrow-compatible table, normally selected from the data SQL relation. Keyword arguments replace fields of the collection, such as id or description; the rest is inherited. For a remote source, metadata is cached and only the payload files belonging to the selected samples are downloaded. Pass overwrite=True to replace an existing TACO output.

source = "https://data.source.coop/major-tom/core-dem/"
dataset = taco.open_dataset(source)
rows = dataset.sql("SELECT * FROM data ORDER BY sample_id LIMIT 10")
taco.export(
    source,
    "core-dem-sample.zip",
    samples=rows,
    id="core-dem-sample",
    description="Ten samples from Core-DEM",
)

Remote reads and exports show download progress in interactive terminals. Writers show their build progress when opened with progress=True.

Examples

Every example is self-contained, uses synthetic data, and writes its output in the current directory.

Spatial and temporal metadata use separate profiles: Spatial for regular spatial grids, ISpatial for irregular footprints, and Temporal for time alone. STAC combines regular spatial + temporal metadata; ISTAC combines irregular spatial + temporal metadata.

Example What it demonstrates
minimal.py Smallest possible single-file dataset
numpy_minimal.py NumPy image and mask assets with a train/test split
change_detection.py Metadata on before/ and after/ folders
sequence.py Variable-length asset sequences
time_series.py Per-observation time and cloud metadata
geospatial.py Compact STAC metadata and derived MajorTOM cells
stac_segmentation.py STAC extensions for regular raster chips, labels, bands, and scaling
oceantaco_istac.py OceanTACO-inspired ISTAC metadata for irregular SWOT swaths and Argo collocations
partitioned.py ZIP partitions and their TACOCAT catalog