reproject

What it is and what it does

Reproject is a Python package for re-gridding astronomical images from one world coordinate system to another—changing pixel resolution, orientation, or coordinate frame. It implements three distinct reprojection techniques: simple interpolation (similar to SWARP), the adaptive anti-aliased algorithm from DeForest (2004), and exact pixel overlap calculation on the celestial sphere (similar to Montage). It also supports reprojection to and from HEALPIX projections via astropy-healpix.

The package depends on a substantial stack of scientific Python libraries (numpy, scipy, astropy, astropy-healpix) and integrates with dask, dask-image, zarr, and fsspec for handling larger datasets. It is actively maintained, requires Python 3.11 or later, and is distributed under a permissive BSD 3-Clause license. Installation involves pre-built wheels for standard platforms, though the compiled dependencies add moderate friction.

Use it for:

• Align multiple astronomical survey images to a common pixel grid and coordinate system before stacking or analysis.
• Convert celestial observations between different projections (e.g., tangent-plane to HEALPIX) for survey-wide processing.
• Resample high-resolution imaging data to match lower-resolution reference frames while preserving flux.
• Prepare multi-wavelength observations from different telescopes for joint analysis by harmonizing their coordinate systems.
• Generate footprint maps showing valid data coverage after reprojection to a target coordinate frame.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Reproject astronomical images between different world coordinate systems and pixel grids using interpolation, adaptive anti-aliasing, or exact pixel overlap methods, with support for HEALPIX projections.

Yes, if you work with astronomical images and need to align them across different coordinate systems or resolutions. The package offers a well-maintained, permissive-licensed solution with multiple reprojection algorithms and HEALPIX support. Medium install friction is acceptable for scientific workflows. No known security vulnerabilities.

Install

reproject on PyPI

pip install reproject

uv add reproject

poetry add reproject

Installing reproject

Before you install

Medium install friction due to 10 runtime dependencies including compiled packages (numpy, scipy, astropy). Pre-built wheels available for current Python versions on major platforms. Active maintenance with recent release.

License in practice

BSD 3-Clause is permissive; you can use, modify, and distribute this package freely in commercial and private projects provided you include the license notice.

Quickstart

pip install reproject

from reproject import reproject_interp
from astropy.io import fits

data, header = fits.getdata('input.fits', header=True)
output_data, footprint = reproject_interp((data, header), target_header)

Requires Python >= 3.11. Compiled dependencies (numpy, scipy, astropy) must be installed; pre-built wheels handle this on standard platforms.

Verify before relying

• Whether the package handles large datasets efficiently with dask and zarr integration beyond basic reprojection.
• Performance characteristics and memory requirements for typical astronomical survey data sizes.
• Specific accuracy or speed trade-offs between the three implemented reprojection algorithms.

Package facts

Evidence: reproject-0.21.0-cp311-abi3-macosx_10_9_x86_64.whl; reproject-0.21.0-cp311-abi3-macosx_11_0_arm64.whl; reproject-0.21.0-cp311-abi3-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl; reproject-0.21.0-cp311-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl; reproject-0.21.0-cp311-abi3-win_amd64.whl

Tags

More Scientific/Engineering packages