Installation

Installation#

Requirements#

scDataset requires Python 3.8 or higher and the following dependencies:

  • torch >= 1.2.0

  • numpy >= 1.17.0

Optional dependencies for specific data formats:

  • anndata - for AnnData support

  • datasets - for HuggingFace Datasets support

Install from PyPI#

The easiest way to install scDataset is from PyPI:

pip install scDataset

This will install the latest stable release along with all required dependencies.

Install from GitHub#

To get the latest development version, install directly from GitHub:

pip install git+https://github.com/Kidara/scDataset.git

Development Installation#

For development, clone the repository and install in editable mode:

git clone https://github.com/Kidara/scDataset.git
cd scDataset
pip install -e .

To install development dependencies:

pip install -e ".[dev]"

Verify Installation#

You can verify your installation by importing the package:

import scdataset
print(scdataset.__version__)

Or run a quick test:

from scdataset import scDataset, Streaming
import numpy as np

# Create test data
data = np.random.randn(100, 50)
dataset = scDataset(data, Streaming(), batch_size=10)

# Test iteration
for batch in dataset:
    print(f"Batch shape: {batch.shape}")
    break

print("Installation successful!")

Troubleshooting#

ImportError: No module named ‘torch’

Make sure PyTorch is installed. Visit pytorch.org for installation instructions.

Performance Issues

For best performance with large datasets, consider installing:

pip install numba  # For faster numerical operations
Contents