MPSKit

MPSKit v0.13.0

Diff since v0.12.6

MPSKit v0.13 features a number of breaking changes that should overall improve clarity and performance. There are two major changes that should be noted:

The derivative operators have seen a major rewrite to unify their calling signatures, and MPOHamiltonian versions now have dedicated structures to more efficiently make use of the particular structure of the MPO. Overall, we expect this to improve performance, especially for short-range hamiltonians this should be noticeable. As part of this refactor, we are now more consistently using below, operator, above to allow for different states in the bra and ket layers.

The conventions for the sector keyword we're somewhat counterintuitive, effectively having to input the dual of what would be expected. This is now changed, and made consistent across the package.

Finally, various small fixes and improvements were made.

Merged pull requests:

• [New] Finite temperature utilities + docs example (#262) (@lkdvos)
• Update mpo.jl (#264) (@dartsushi)
• Refactor VUMPS and VOMPS to avoid boxed variables (#265) (@lkdvos)
• Fix finite temperature example (#266) (@lkdvos)
• Consistent environments signature: below, operator, above argument order (#268) (@leburgel)
• Refactor derivatives and add JordanMPOTensor (#270) (@lkdvos)
• Change sector convention for excitations (#271) (@lkdvos)
• Improve quasiparticle type stability (#272) (@lkdvos)
• Fix add_physical_charge for JordanMPOTensor (#273) (@lkdvos)
• Add tests for fermionic periodic boundary conditions (#274) (@lkdvos)
• fix typo in localupdate_step for VOMPS algorithm (#275) (@VictorVanthilt)
• Refactor derivative names (#276) (@lkdvos)
• Fermionic test no longer assumes h = permute(h, ((2, 1), (4, 3))) (#277) (@lkdvos)
• Fix time dependent integral of TDVP and TDVP2 (#279) (@GiggleLiu)
• Update v0.13.0 (#280) (@lkdvos)

Closed issues:

• Unexpected behavior for sector keyword argument in e.g. exact_diagonalization (#260)
• A bug? (#267)