Tempo requires only a computer with enough RAM to support in-memory operations. For most datasets, >=8GB RAM should be sufficient.
Tempo has been tested on macOS 10.14.5 (Mojave), 12.2.1 (Monterey), and CentOS Linux 7.
Tempo requires python >= 3.8 and depends on the following python packages:
- anndata
- numpy
- pandas
- scanpy>=1.6
- scikit-image
- scikit-learn
- scipy
- statsmodels
- torchaudio
- torchvision
- tqdm
- pytorch>=1.9.0
Tempo requires mini-conda for installation. For information on installing mini-conda for your operating system, please view https://docs.conda.io/en/latest/miniconda.html.
After installing mini-conda, run the following commands to install Tempo:
- git clone https://github.com/bauerbach95/tempo
- cd tempo
- source install.sh
Installation should take less than 5 minutes. After installing, you can activate the conda environment containing the installed Tempo package: conda activate tempo
To test if Tempo works properly, run the run_test.py file using using the activated environment: python run_test.py
Running the test should take less than 5 minutes. Successful completion of the test will yield a message "SUCCESSFULLY FINISHED".
To deactivate the environment: conda deactivate
Tutorials on how to run Tempo (and information on inputs / outputs) can be viewed in the tutorial folder of the repository.