HDF5Array performance
Introduction | Install and load the required packages | The test datasets | Sparse vs dense representation | TENxMatrix vs HDF5Matrix objects | Bring the sparse dataset in R | Bring the dense dataset in R | Create the test datasets | Block-processed normalization and PCA | Code used for normalization and PCA | Block processing and block size | Monitoring memory usage | Normalization and PCA of the 27,998 x 12,500 test dataset | Normalization | TENxMatrix (sparse) | HDF5Matrix (dense) | HDF5Matrix as sparse | On-disk realization of the normalized datasets | PCA | Comprehensive timings obtained on various machines | Timings for DELL XPS 15 laptop | Timings for Supermicro SuperServer 1029GQ-TRT | Timings for Apple Silicon Mac Pro | Timings for Intel Mac Pro | Discussion | TENxMatrix (sparse) vs HDF5Matrix (dense) | Hybrid approach | A note about memory usage | Main takeaways | Session information