TorchDR is a dimensionality reduction (DR) library designed to offer a unified interface for a range of widely used DR methods, including t-SNE, UMAP, LargeVis, and others. The library achieves high computational efficiency by leveraging GPU acceleration, making it especially well-suited for handling large, high-dimensional data such as single-cell datasets.
TorchDR is a dimensionality reduction (DR) library designed to offer a unified interface for a range of widely used DR methods, including t-SNE, UMAP, LargeVis, and others. The library achieves high computational efficiency by leveraging GPU acceleration, making it especially well-suited for handling large, high-dimensional data such as single-cell datasets.