Release notes

v 1.4.0

December 14, 2023

Implement the block format so that Dcut can be any number

v 1.3.5

December 6, 2023

Fix the bug that says Grassmann tensor is not even when doing SVD, Eig, or Hconjugate

v 1.3.4

November 6, 2023

Added the sympy dependency.
Correction of the initial tensor generation.

v 1.3.3

October 25, 2023

Added the operation __truediv__ to dense and sparse.

v 1.3.2

October 25, 2023

Added the operation __pos__ to dense, sparse, and grassmann_number.

v 1.3.1

October 25, 2023

Added the operations __neg__, __radd__, __sub__ (redefined as +(-1)*other), and __rsub__, to dense and sparse.
Added missing exception handlings for __mul__ in dense and sparse.

v 1.3.0

October 24, 2023

Added the arith module, which can be used to generate the initial tensor from the action directly in Python. Symbolic computation supported.
Modified the gauge2d module to make use of arith instead of the imported code from mathematica. The original code contains some bugs, which are now fixed.
Changed the function join_legs() so that the encoding of the result is the same as the original object in most cases.

v 1.2.3

August 31, 2023

This is the version described in the preprint