Installing MRI-NUFFT

mri-nufft is available on PyPi

Tip

TLDR: If you have a GPU you probably want to install MRI-NUFFT with one of the following options:

• uv add mri-nufft[cufinufft] generally faster, but memory hungry

• uv add mri-nufft[gpunufft] memory efficient, and still fast.

• uv add mri-nufft[finufft] for non-cuda setup (CPU only) setup.

(the regular pip install or uv pip install would work as well).

Then, use the get_operator(backend=<your backend>, ... ) to initialize your MRI-NUFFT operator.

You are ready to use MRI-NUFFT ! Go check the Examples for more information.

pip install mri-nufft

However, if you want to use some specific backend or develop on mri-nufft, you can install it with extra dependencies. notably extra, io, and autodiff

uv add mri-nufft[extra,io,autodiff]

Tip

using uv will ensure that the correction version of pytorch (with CUDA) is installed. for more info see `this < https://docs.astral.sh/uv/guides/integration/pytorch/>`__

Using uv

If you are using uv as your package installer you will need to do

uv add mri-nufft[extra,io,autodiff]

Development Version

If you want to modify the mri-nufft code base:

git clone https://github.com:mind-inria/mri-nufft
pip install -e ./mri-nufft[dev,doc,extra,io,autodiff,tests,cufinufft,gpunufft,finufft]

or using uv:

git clone https://github.com:mind-inria/mri-nufft
uv venv
uv sync --all-extras --no-build-isolation --no-extra <backend-you-dont-need>

Contributing to MRI-NUFFT

Check our CONTRIBUTING.md