v0.14.8

TensorKit v0.14.8

Diff since v0.14.7

Merged pull requests:

• Refactor repacking of transformer structure (#254) (@lkdvos)
• [Fix] Transformers for non-abelian non-selfadjoint sectors (#256) (@lkdvos)

Closed issues:

• Contraction error after v0.14.7 (#255)

v0.14.7

TensorKit v0.14.7

Diff since v0.14.6

This patch release contains some non-breaking changes in the form of important performance reworks that are particularly relevant to non-Abelian symmetries in the context of tensors with a larger number of indices.
The backend for index manipulations for these kinds of tensors should be noticeably faster.
For those that wish the experiment, you can also now configure the multithreading settings for these operations through set_num_transformer_threads(n).
Additionally, the system for keeping the cached data has been refactored and cleaned up, so it should be slightly easier to interact with that and print information about it.
Finally, various minor convenience and docs improvements have been carried out.

Merged pull requests:

• make all caches behave the same (#244) (@Jutho)
• Default spacetype and sectortype implementations in type domain (#248) (@ogauthe)
• DiagonalTensorMap performance and convenience specializations (#249) (@lkdvos)
• small type stability fix (#250) (@Jutho)
• [Performance] TreeTransformer refactor + multithreading (#251) (@lkdvos)
• Add color inversion for docs images in dark themes (#253) (@leburgel)

Closed issues:

• Dark theme of website "https://jutho.github.io/TensorKit.jl/stable/" (#128)
• Can the category VecG be supported? (#227)
• About functions on braided sector (#246)

v0.14.6

TensorKit v0.14.6

Diff since v0.14.5

This patch release features mostly small fixes and improvements.
Most notable are the addition of rrule support for inv, DiagonalTensorMap and flip, and the ability to use in-place versions of the special-purpose constructors like isomorphism!, unitary! etc.

Merged pull requests:

• Add complex(::BraidingTensor) specialization (#226) (@lkdvos)
• Replace isequal with == for FusionTree (#228) (@lkdvos)
• Correct argument names in planartrace function (#229) (@lkdvos)
• correct typo in docs (#231) (@VictorVanthilt)
• Rewrite special-purpose constructors to use in-place operations (#232) (@lkdvos)
• Add benchmark suite (#233) (@lkdvos)
• Add rrule for DiagonalTensorMap and include tests (#234) (@lkdvos)
• Add rrule for inv and corresponding tests (#236) (@lkdvos)
• Improve error handling for tensor functions (#238) (@lkdvos)
• Allow twisting of trivial sectors without restrictions (#240) (@lkdvos)
• add rrule for flip (#241) (@sanderdemeyer)

Closed issues:

• Error generates an error (#237)
• twist for NoBraiding sector types where the braiding is unambiguous (#239)

v0.14.5

TensorKit v0.14.5

Diff since v0.14.4

Merged pull requests:

• Add rank and cond (#220) (@lkdvos)
• Fix QR and LQ pullback for rank 0 blocks (#223) (@lkdvos)
• insertunit and removeunit using Val (#224) (@Jutho)
• Replace c == one(I) with isone(c) (#225) (@lkdvos)

Closed issues:

• MethodError in svd_pullback! (#219)
• @constinferred fails for insertleftunit and insertrightunit of HomSpaces for particular Sectors (#221)
• MethodError in qr_pullback! (#222)

v0.14.4

TensorKit v0.14.4

Diff since v0.14.3

Merged pull requests:

• Add dedicated BlockIterators (#206) (@lkdvos)
• DiagonalTensorMap constructor rrule (#208) (@ebelnikola)
• Fix rrules of TensorOperations with DiagonalTensorMap (#210) (@lkdvos)
• Add DiagonalTensorMap constructors and converters (#212) (@Yue-Zhengyuan)
• Add rrule for matrix exponential (#214) (@lkdvos)
• Nightly tests don't count as failed CI (#215) (@lkdvos)
• Add rrule for twist (#217) (@lkdvos)
• Attempt to use default arch in actions to avoid complaints (#218) (@lkdvos)

Closed issues:

• An AD issue in contractions with diagonal tensors (#209)
• rrule for exponential of TensorMap (#213)
• Autodiff for twist? (#216)

v0.14.3

TensorKit v0.14.3

Diff since v0.14.2

Merged pull requests:

• fixed typo (#199) (@adamnemecek)
• rrule for copy_oftype, permutedcopy_oftype (#202) (@lkdvos)
• Fix leftorth! and rightorth! with empty blocks (#205) (@Jutho)
• add unthunk in rrules of eig! eigh! leftorth! rightorth! (#207) (@tangwei94)

Closed issues:

• repartition unexpected behaviour (#200)
• AD of SVD question (#201)
• a small issue of leftorth! (#204)

v0.14.2

TensorKit v0.14.2

Diff since v0.14.1

Merged pull requests:

• Fix similar (#195) (@Jutho)
• fix ad type stability (#198) (@Jutho)

Closed issues:

• libblas not available on nightly (#65)
• The initial functions TensorMap(zeros) and similar appear to be time-consuming for large tensors. (#147)
• Undefined constructor. (#153)
• similar does not work if storagetype is a PtrArray (#194)
• rrules have type instabilities (#197)

v0.14.1

TensorKit v0.14.1

Diff since v0.14.0

• Fix an issue with fusiontree iterators for symmetries with multiple fusion.

Merged pull requests:

• Fix bug in fusion tree iterator for multiple fusion case (#193) (@Jutho)

Closed issues:

• Fusiontree iterator with multiplicities (#192)

v0.14.0

TensorKit v0.14.0

Diff since v0.13.2

• Various index functionalities have been added: flip can be used to change arrows, insertleftunit, insertrightunit and removeunit will manipulate trivial indices.

Breaking changes

• The DiagonalTensorMap is now the default output type for the diagonal singular and eigenvalue tensors returned by tsvd, eig and related functions
• There is a small change in the truncation of singular values for non-abelian symmetries, where now it is guaranteed that smaller singular values are removed first

Merged pull requests:

• Add and test rrules for real and imag (#183) (@lkdvos)
• Constructor fixes (#185) (@Jutho)
• insertleftunit, insertrightunit and removeunit (#187) (@lkdvos)
• change truncation strategy to always select smallest schmidt coefficient (#188) (@Jutho)
• add 🐬 functionality (#189) (@Jutho)
• use diagonaltensormap (#190) (@Jutho)
• some cleanup for v0.14 (#191) (@Jutho)

Closed issues:

• svdvals and eigvals (#34)
• How to apply gradient of gradient? (#175)
• convert to floating point in matrix factorisations (#181)
• Fail to extract block by specifying sector of each index for (N,0) or (0,N) tensors (#182)
• Why using weighted singular values for the truncation? (#184)
• Partial convert(Array, TensorMap) (#186)

v0.13.2

TensorKit v0.13.2

Diff since v0.13.1

• Add functionality for working with DiagonalTensorMap (currently not yet used)
• Various fixes and improvements

Merged pull requests:

• WIP: implementation of DiagonalTensorMap (#174) (@Jutho)
• CompatHelper: bump compat for VectorInterface to 0.5, (keep existing compat) (#176) (@github-actions[bot])
• Bump codecov/codecov-action from 4 to 5 (#177) (@dependabot[bot])
• Braidingtensor improvements (#179) (@lkdvos)
• Fix converting empty tensors to array (#180) (@lkdvos)

Closed issues:

• convert to Array fails for empty TensorMap (#178)