Installation

SpaHDmap can be installed using pip or directly from the source code and supports for both Linux and Windows operating systems. We have tested the installation of Version 0.0.3 on Linux and Windows with a virtual environment using Python 3.9.27. It takes around 5 minutes to install SpaHDmap on a standard laptop (most of the time is spent on downloading the dependencies). Here are the different methods to install SpaHDmap:

Using pip

The easiest way to install SpaHDmap is using pip:

pip install SpaHDmap

This will install the latest stable version of SpaHDmap along with all its dependencies.

For large dataset, we recommend installing SpaHDmap with annoy for faster nearest neighbor search:

pip install SpaHDmap[annoy]

From Source

To install SpaHDmap from source, first clone the repository:

git clone https://github.com/sldyns/SpaHDmap.git

Then, install the package using pip:

pip install -e .

This will install SpaHDmap in editable mode, which is useful if you want to modify the source code.

Requirements

SpaHDmap requires Python 3.9 or later. The main dependencies are:

  • numpy>=1.22

  • cython>=0.29.24

  • torch>=1.12

  • scikit-learn>=1.0.1

  • squidpy>=1.6.0

  • matplotlib>=3.4.3

  • tqdm>=4.62.3

  • h5py>=3.10.0

  • scanpy>=1.9.7

  • anndata>=0.10.0

  • pandas>=2.2.0

  • scikit-image>=0.23.0

  • opencv-python>=4.10.0

  • imagecodecs>=2023.1.23

  • scipy>=1.13.1

  • seaborn>=0.13.1

  • annoy>=1.17.0 (optional, for approximate nearest neighbor search)

These dependencies will be automatically installed when you install SpaHDmap using pip.

GPU Support

SpaHDmap can leverage GPU acceleration for faster computation. To use GPU support, make sure you have a CUDA-compatible GPU and the appropriate CUDA toolkit installed. Then, install the GPU version of PyTorch:

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu113

Replace cu113 with the appropriate CUDA version for your system.

Verifying Installation

After installation, you can verify that SpaHDmap is correctly installed by running:

import SpaHDmap as hdmap
print(hdmap.__version__)

This should print the version number of SpaHDmap without any errors.

Troubleshooting

If you encounter any issues during installation, please check the following:

  1. Ensure you have the latest version of pip:

    pip install --upgrade pip

  2. If you’re using a virtual environment, make sure it’s activated.

  3. On Windows, you might need to install Microsoft Visual C++ Build Tools.

If you still face problems, please open an issue on the SpaHDmap GitHub repository.