Installation

This package is tested on Linux/Windows/Mac OS X. Pre-built binaries should be available for platform-specific dependencies. No manual compilation should be needed.

For end users who simply want to use or run scripts depending on ASPIRE, installing the aspire package from PyPI is sufficient.

Note

Installing the package installs ASPIRE to the site-packages folder of your active environment. This is only desirable if you are not going to be doing any development on ASPIRE, and only intend to run scripts that depend on the ASPIRE package.

For those who wish to develop, we recommend starting with the instructions on our README (copied below). Additionally some more advanced instructions are provided here for installing with software and hardware optimizations. Although not explicitly required, For developers and users not confident in software management the use of conda is strongly encouraged.

Install Conda

To follow the suggested installation, you will need to install Conda for Python3, either Anaconda or Miniconda, click on the right distribution to view Conda’s installation instructions.

Note

If you’re not sure which distribution is right for you, go with Miniconda

Note

For Apple silicon to use the osx-arm platform, patching and building some dependencies from source is currently required. The Intel osx-64 install is still preferred even for Apple silicon users, otherwise notes are provided.

Getting Started - Installation

Python 3.8 is used as an example, but the same procedure should work for any of our supported Python versions 3.8-3.11. Below we pip install the aspire package using the -e flag to install the project in editable mode. The ".[dev]" command installs aspire from the local path with additional development tools such as pytest and Jupyter Notebook. See the pip documentation for more details on using pip install.

# Clone the code
git clone https://github.com/ComputationalCryoEM/ASPIRE-Python.git
cd ASPIRE-Python

# Create a fresh environment
conda create --name aspire python=3.8 pip

# Enable the environment
conda activate aspire

# Install the ``aspire`` package from the checked out code
# with the additional ``dev`` extension.
pip install -e ".[dev]"

Note

Required dependent packages supporting Python 3.11 are currently only supported on Linux.

Test the package

Make sure all unit tests run correctly by doing:

pytest

Tests currently take around 15 minutes to run, but this depends on your specific machine’s resources and configuration.

Optimized Numerical Backends

For advanced users, conda provides optimized numerical backends that offer significant performance improvements on appropriate machines. The backends accelerate the performance of numpy, scipy, and scikit packages. ASPIRE ships several environment*.yml files which define tested package versions along with these optimized numerical installations.

The default environment-default.yml does not force a specific backend, instead relying on conda to select something reasonable. In the case of an Intel machine, the default conda install will automatically install some optimizations for you. However, these files can be used to specify a specific setup or as the basis for your own customized conda environment.

Suggested Conda Environments

Architecture	Recommended Environment File
Default	environment-default.yml
Intel x86_64	environment-intel.yml
AMD x86_64	environment-openblas.yml
Apple M1	environment-accelerate.yml

Using any of these environments follows the same pattern outlined below. As an example to specify using the accelerate backend on an M1 laptop:

cd ASPIRE-Python
conda env create -f environment-accelerate.yml --name aspire_acc
conda activate aspire_acc
pip install -e ".[dev]"

Installing GPU Extensions

ASPIRE does support GPUs, depending on several external packages. The collection of GPU extensions can be installed using pip. Extensions are grouped based on CUDA versions. To find the CUDA driver version, run nvidia-smi on the intended system.

CUDA GPU Extension Versions

CUDA Version	ASPIRE Extension
10.2	gpu-102
11.0	gpu-110
11.1	gpu-111
>=11.2	gpu-11x
>=12	gpu-12x

For example, if you have CUDA 12.3 installed on your system, the command below would install GPU packages required for ASPIRE.

# From a local git repo
pip install -e ".[gpu-12x]"

# From PyPI
pip install "aspire[gpu-12x]"

By default if the required GPU extensions are correctly installed, ASPIRE should automatically begin using the GPU for select components (such as those using nufft).

Because GPU extensions depend on several third party packages and libraries, we can only offer limited support if one of the packages has a problem on your system.

Generating Documentation

Sphinx Documentation of the source (a local copy of what you’re looking at right now) can be generated by using the following commands from the root of the code repository.

ASPIRE has both traditional documentation and a gallery of tutorial scripts. To make only the documentation run make html-noplot. The make html command makes the traditonal documentation then runs and renders the gallery/tutorials examples, which takes several minutes.

cd docs

# Parse the code in ``src``
sphinx-apidoc -f -o ./source ../src -H Modules

make html-noplot  # Generate only documentation
# or
make html         # Generate documentation and gallery examples

# To remove any documentation build artifacts
make distclean

The resulting html files can be found at docs/build/html.