Change LGST Circuit Storage with FPR (Color Boxplot Fix)

[coreyostrove]

This changes the way the LGST circuits are stored when using FPR. Previously these were stripped from the plaquettes and any needed circuits were added back in in _additional_circuits. Now we just don't remove those for L=1 if we shouldn't. This addresses a bug reported by @pcwysoc that was present in the color box plots where the result of the old behavior was that these LGST circuits weren't rendered with the plaquettes as expected.

[sserita]

One point of discussion.

[sserita]

Just a general note that I've noticed several of these OrderedDict -> dict changes in recent PRs. It's not a problem, but just wanted to link it back to our discussion in #427 where we wanted to annotate dicts where order is explicitly important (e.g. where keys/values are iterated over with an expected order later).
I only bring this up here since "keeping consistent ordering" is part of the comment for this dict instantiation - is this one such place where order matters, and we should finalize our decision from #427?

[coreyostrove]

Yup, good catch. I have been adding the annotations where relevant (which has actually been pretty uncommon, it turns out), but forgot it seems. I'll add these back in when I get a chance.

…

On Mon, Mar 31, 2025 at 4:53 PM Stefan Seritan ***@***.***> wrote: ***@***.**** commented on this pull request. One point of discussion. ------------------------------ In pygsti/circuits/gstcircuits.py <#561 (comment)>: > @@ -528,7 +527,7 @@ def add_to_plaquettes(pkey_dict, plaquette_dict, base_circuit, maxlen, germ, pow if nest: #keep track of running quantities used to build circuit structures running_plaquette_keys = {} # base-circuit => (maxlength, germ) key for final plaquette dict - running_plaquettes = _collections.OrderedDict() # keep consistent ordering in produced circuit list. + running_plaquettes = dict() # keep consistent ordering in produced circuit list. Just a general note that I've noticed several of these OrderedDict -> dict changes in recent PRs. It's not a problem, but just wanted to link it back to our discussion in #427 <#427> where we wanted to annotate dicts where order is explicitly important (e.g. where keys/values are iterated over with an expected order later). I only bring this up here since "keeping consistent ordering" is part of the comment for this dict instantiation - is this one such place where order matters, and we should finalize our decision from #427 <#427>? — Reply to this email directly, view it on GitHub <#561 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ASY6RPRMCNOMHTRJINAOGYD2XHBODAVCNFSM6AAAAABZZC6EDKVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDOMZQGY4TCNRRGY> . You are receiving this because you authored the thread.Message ID: ***@***.***>