Bug fixes, broadcasting, matricize, and display improvements#151

Merged

mtfishman merged 18 commits into

mainfrom

mf/demo-readiness

Apr 15, 2026

Commits on Apr 15, 2026

fix: handle β=0 in scale! and remove iszero guard in permutedimsopadd!

mtfishman
and
claude
committed
feat: rename AbelianGradedStyle → GradedStyle and enable FusedGradedMatrix broadcasting

mtfishman
and
claude
committed
fix: use compact SectorRange display instead of TensorKitSectors label()

mtfishman
and
claude
committed
feat: AbstractSectorArray displays as Kronecker structure (sector ⊗ data)

mtfishman
and
claude
committed
feat: FusedGradedMatrix Pair constructor

mtfishman
and
claude
committed
feat: block-structured display for AbelianGradedArray and FusedGradedMatrix

mtfishman
and
claude
committed
feat: block-structured display for AbelianGradedArray and FusedGradedMatrix

mtfishman
and
claude
committed
feat: printing refinements for graded/sector types

mtfishman
and
claude
committed
feat: round 2 printing refinements

mtfishman
and
claude
committed
feat: FusedGradedMatrix text/plain display shows per-axis dims

mtfishman
and
claude
committed
refactor: simplify matricize padding, remove _mul_mismatched and _pad_fused

mtfishman
and
claude
committed
refactor(pad_to_canonical_duals): use block view + `sectoraxes` for remap

mtfishman
and
claude
committed
rename: pad_to_canonical_duals → insert_missing_sectors + preconditions

mtfishman
and
claude
committed
refactor: extract `delete_missing_sectors` into fusion.jl, generic over AbstractGradedMatrix

mtfishman
and
claude
committed
refactor: route `insert_missing_sectors` and `delete_missing_sectors` preconditions through `check_input`

mtfishman
and
claude
committed
Drop redundant zero! in AbelianGradedArray(m) constructor, annotate β=0 fix

mtfishman
and
claude
committed
Audit `zero!` uses: drop two redundant calls, annotate the rest

mtfishman
and
claude
committed
refactor(delete_missing_sectors): simplify checks via per-axis set ops

mtfishman
and
claude
committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Bug fixes, broadcasting, matricize, and display improvements#151

Bug fixes, broadcasting, matricize, and display improvements#151
mtfishman merged 18 commits into
mainfrom
mf/demo-readiness

Commits on Apr 15, 2026

fix: handle β=0 in scale! and remove iszero guard in permutedimsopadd!

feat: rename AbelianGradedStyle → GradedStyle and enable FusedGradedMatrix broadcasting

fix: use compact SectorRange display instead of TensorKitSectors label()

feat: AbstractSectorArray displays as Kronecker structure (sector ⊗ data)

feat: FusedGradedMatrix Pair constructor

feat: block-structured display for AbelianGradedArray and FusedGradedMatrix

feat: block-structured display for AbelianGradedArray and FusedGradedMatrix

feat: printing refinements for graded/sector types

feat: round 2 printing refinements

feat: FusedGradedMatrix text/plain display shows per-axis dims

refactor: simplify matricize padding, remove _mul_mismatched and _pad_fused

refactor(pad_to_canonical_duals): use block view + `sectoraxes` for remap

rename: pad_to_canonical_duals → insert_missing_sectors + preconditions

refactor: extract `delete_missing_sectors` into fusion.jl, generic over AbstractGradedMatrix

refactor: route `insert_missing_sectors` and `delete_missing_sectors` preconditions through `check_input`

Drop redundant zero! in AbelianGradedArray(m) constructor, annotate β=0 fix

Audit `zero!` uses: drop two redundant calls, annotate the rest

refactor(delete_missing_sectors): simplify checks via per-axis set ops