TidyTensor: Utilities for multidimensional arrays as named hierarchical structures
Shawn T. O’Neil · The Journal of Open Source Software · 2021
Deep learning applications commonly employ the use of tensors (multidimensional arrays) for data storage and manipulation.In addition to being space efficient tensors provide a common framework for, the engine of deep neural network training (Baydin et al., 2018).TidyTensor is an R package for inspecting and manipulating tensors.It provides an improved print() function for summarizing structure, named tensors, conversion to data frames, and high-level manipulation functions.Designed to complement keras package (or other packages that utilize native R arrays), functionality is layered on top of base R types.