TODO List

This lists all the .. todo blocks in all docstrings. For additional outstanding tasks, features, and bugs, see the issues on github.

Todo

We should add a method to do fermion matrix-vector multiply, perhaps as __call__. The programming puzzle is that when \(\vec{h}=0\) or \(\vec{h}\parallel \hat{z}\) we can split \(\mathbb{d}\) into two independent pieces that each operate on spacetime vectors, while for generic \(\vec{h}\) we need one matrix that operates on a spacetime ⊗ spin vector.

(The original entry is located in /home/docs/checkouts/readthedocs.org/user_builds/two-dimensional-gasses/checkouts/latest/python/tdg/tdg/fermionMatrix.py:docstring of tdg.fermionMatrix.FermionMatrix, line 49.)

Todo

Leveraging this makes some code much slower. For example,

>>> UU = torch.stack(
...     tuple(Action.FermionMatrix.UU_tensor(cfg) for cfg in configurations)
...     ).permute(0,1,3,2,4).reshape(-1,
...                                  2*Action.FermionMatrix.Spacetime.Lattice.sites,
...                                  2*Action.FermionMatrix.Spacetime.Lattice.sites)

is much faster to subsequently manipulate than

>>> UU = torch.stack(
...     tuple(Action.FermionMatrix.UU(cfg) for cfg in configuration))

which is something that ought to be understood / addressed. Perhaps it is autograd- or computational-graph-related?

Is it possible to construct this code to distribute over a whole ensemble of auxiliary fields? This perhaps should inform the way we code other observables.

(The original entry is located in /home/docs/checkouts/readthedocs.org/user_builds/two-dimensional-gasses/checkouts/latest/python/tdg/tdg/fermionMatrix.py:docstring of tdg.fermionMatrix.FermionMatrix.UU, line 7.)

Todo

The starting weight should automatically be read in; currently not.

(The original entry is located in /home/docs/checkouts/readthedocs.org/user_builds/two-dimensional-gasses/checkouts/latest/python/tdg/tdg/ensemble.py:docstring of tdg.ensemble.GrandCanonical.continue_from, line 11.)

Todo

In fact, it is SO NOISY that it has not been checked for correctness by comparing with an exact Trotterized two-body calcuation.

(The original entry is located in /home/docs/checkouts/readthedocs.org/user_builds/two-dimensional-gasses/checkouts/latest/python/tdg/tdg/observable/contact.py:docstring of tdg.observable.contact.Contact_bosonic, line 4.)