v0.3.49

ITensors v0.3.49

Diff since v0.3.48

Merged pull requests:

• [ITensors] Tests for general fermionic H (#1200) (@emstoudenmire)
• [NDTensors] Unwrap module for dispatching on unwrapped types (#1220) (@kmp5VT)
• [NDTensors] Get more Array storage functionality working (#1222) (@mtfishman)
• CompatHelper: add new compat entry for Statistics at version 1 for package ITensorUnicodePlots, (keep existing compat) (#1223) (@github-actions[bot])
• CompatHelper: add new compat entry for Statistics at version 1 for package ITensorVisualizationBase, (keep existing compat) (#1224) (@github-actions[bot])
• [NDTensors] Add DiagonalArrays submodule (#1225) (@mtfishman)
• [NDTensors] DiagonalArray tensor operations (#1226) (@mtfishman)
• [ITensorGaussianMPS] Fix keyword argument forwarding issue (#1228) (@mtfishman)

Closed issues:

• [NDTensors] [BUG] Many basic tensor operations fail with CUDA extension due to LoadError: Scalar indexing is disallowed. (#1193)

v0.3.48

ITensors v0.3.48

Diff since v0.3.47

Merged pull requests:

• [ITensors] Optimize directsum again (#1221) (@mtfishman)

v0.3.47

ITensors v0.3.47

Diff since v0.3.46

Merged pull requests:

• [ITensors] Customisable Arrow directions in SVD (#1197) (@JoeyT1994)
• Fix behavior of factorize mindim (#1214) (@ryanlevy)
• [NDTensors] Get more block sparse operations working on GPU (#1215) (@mtfishman)
• [NDTensors] Avoid more scalar indexing operations in block sparse GPU code (#1217) (@mtfishman)
• [NDTensors][Bug] Fix CPU performance issue (#1218) (@kmp5VT)

Closed issues:

• [ITensors] [BUG] Inconsistent Behavior of mindim (#1207)

v0.3.46

ITensors v0.3.46

Diff since v0.3.45

Merged pull requests:

• Fix fermion cases in blocksparse SVD (#1212) (@emstoudenmire)

v0.3.45

ITensors v0.3.45

Diff since v0.3.44

Merged pull requests:

• [NDTensors][NDTensorsCUDAExt] Improve performance of GPU backends (#1194) (@kmp5VT)

v0.3.44

ITensors v0.3.44

Diff since v0.3.43

Merged pull requests:

• [NDTensors] Add SmallVectors submodule (#1202) (@mtfishman)
• [NDTensors] Add SortedSets and TagSets (#1204) (@mtfishman)
• [NDTensors] BlockSparseArrays prototype (#1205) (@mtfishman)
• [NDTensor] Allow general objects are storage backends for Tensor (#1206) (@mtfishman)
• [ITensors] Fix checkflux for QN ITensor with no blocks (#1210) (@emstoudenmire)
• [NDTensors] Fix bugs in SortedSets and SmallVectors (#1211) (@mtfishman)

Closed issues:

• [NDTensors] [BUG] A * B contraction of NDTensors (#1201)
• [ITensors] [BUG] QN correlation matrix not working since #1178 (#1209)

v0.3.43

ITensors v0.3.43

Diff since v0.3.42

Merged pull requests:

• CompatHelper: bump compat for HDF5 to 0.17, (keep existing compat) (#1189) (@github-actions[bot])
• CompatHelper: bump compat for HDF5 to 0.17 for package ITensorGPU, (keep existing compat) (#1190) (@github-actions[bot])
• CompatHelper: bump compat for HDF5 to 0.17 for package NDTensors, (keep existing compat) (#1191) (@github-actions[bot])
• [NDTensors] Improve contract performance for wrapped arrays (#1192) (@mtfishman)
• [ITensors] Use QR in QN factorize (#1195) (@mtfishman)

Closed issues:

• [ITensors] [BUG] DMRG for excited states return wrong state (#1187)

v0.3.42

ITensors v0.3.42

Diff since v0.3.41

Closed issues:

• [ITensors] [BUG] Computing correlation matrix with zero diagonal correlations on a QN MPS (#1179)
• [ITensors] [BUG] ITensors does not compile (#1181)

Merged pull requests:

• [ITensors] [Docs] θ as parameter name in Rz and CRz gates rather than ϕ (#1180) (@terasakisatoshi)
• [ITensors] Improve type stability of svdMPO and qn_svdMPO (#1183) (@terasakisatoshi)
• [ITensors] Optimize directsum (#1185) (@mtfishman)

v0.3.41

ITensors v0.3.41

Diff since v0.3.40

Merged pull requests:

• [ITensors] Fix some fermionic SVD corner cases (#1174) (@emstoudenmire)
• [ITensors] Fix bugs in QN directsum and fermionic SVD (#1178) (@mtfishman)

v0.3.40

ITensors v0.3.40

Diff since v0.3.39

Closed issues:

• [ITensors] [BUG] Precompile Fails (Can't find get_parameters) (#1177)

Merged pull requests:

• [NDTensors] Remove sort(Tuple) in favor of version in Compat (#1169) (@mtfishman)
• [NDTensors] Rename package extensions to follow standard conventions (#1170) (@mtfishman)
• [ITensors] Index filtering docstring improvements (#1171) (@emstoudenmire)
• [NDTensors] Define blockoffsets for Empty tensors/storage (#1172) (@mtfishman)
• [NDTensors] Fix zero for EmptyStorage (#1173) (@mtfishman)
• [ITensors] Allow optionally passing desired output indices of directsum (#1176) (@mtfishman)