Fix noise term crash in region_update!

[tipfom]

Description

This small change fixes the undefined variable encountered when supplying the noise variable to any function using region_update!, e.g., tdvp.

Fixes #130

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 1 line in your changes missing coverage. Please review.

Project coverage is 90.28%. Comparing base (ea26fcd) to head (b21d2c6).

Files with missing lines	Patch %	Lines
src/solvers/sweep_update.jl	0.00%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #146   +/-   ##
=======================================
  Coverage   90.28%   90.28%           
=======================================
  Files          54       54           
  Lines        3583     3583           
=======================================
  Hits         3235     3235           
  Misses        348      348           

Flag	Coverage Δ
docs	18.44% <0.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[mtfishman]

@emstoudenmire based on your comment in #130, should we just disable this instead of fixing it?

[emstoudenmire]

Yes, personally I'd view the idea of combining noise terms / subspace expansions with TDVP as a kind of experimental concept (even at a algorithm research level) so we should disable it here. Even if we do enable it, we'd do so with a different formulation of the subspace expansion that only adds in orthogonal vectors, which is not what the ITensorMPS 'noise' approach currently does.

[mtfishman]

Yes, personally I'd view the idea of combining noise terms / subspace expansions with TDVP as a kind of experimental concept (even at a algorithm research level) so we should disable it here. Even if we do enable it, we'd do so with a different formulation of the subspace expansion that only adds in orthogonal vectors, which is not what the ITensorMPS 'noise' approach currently does.

Agreed. I think if you are doing imaginary time evolution, this formulation of subspace expansion through the noise term could be sufficient, but as you say for real time evolution it is probably not correct, and allowing users to use it is misleading in those cases.