Copyright 2025 National Technology & Engineering Solutions of Sandia, LLC (NTESS). Under the terms of Contract DE-NA0003525 with NTESS, the U.S. Government retains certain rights in this software. 

Welcome to pyttb, a refactor of the Tensor Toolbox for MATLAB in Python.

This package contains data classes and methods for manipulating dense, sparse, and structured tensors, along with algorithms for computing low-rank tensor decompositions:

• Data Classes: tensor, sptensor, ktensor, ttensor, tenmat, sptenmat, sumtensor
• Algorithms: cp_als, cp_apr, gcp_opt, hosvd, tucker_als

python3 -m pip install pyttb 

>>> import pyttb as ttb >>> X = ttb.tenrand((2,2,2)) >>> type(X) <class 'pyttb.tensor.tensor'> >>> M = ttb.cp_als(X, rank=1) CP_ALS: Iter 0: f = 7.367245e-01 f-delta = 7.4e-01 Iter 1: f = 7.503069e-01 f-delta = 1.4e-02 Iter 2: f = 7.508240e-01 f-delta = 5.2e-04 Iter 3: f = 7.508253e-01 f-delta = 1.3e-06 Final f = 7.508253e-01

For historical reasons we use Fortran memory layouts, where numpy by default uses C. This is relevant for indexing. In the future we hope to extend support for both.

>>> import numpy as np >>> c_order = np.arange(8).reshape((2,2,2)) >>> f_order = np.arange(8).reshape((2,2,2), order="F") >>> print(c_order[0,1,1]) 3 >>> print(f_order[0,1,1]) 6

• Documentation
• Tutorials
• Info for users coming from MATLAB
• Learn about tensor decompositions: tensor paper, tensor book

• Report a bug
• Guide for contributors
• List of contributors

If you use pyttb in your work, please cite it using the citation info here.