Skip to content

doc: surface numerical ranking for 3D NS optimization (#86) #92

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
dionhaefner merged 3 commits into from
Jul 9, 2026
Merged

doc: surface numerical ranking for 3D NS optimization (#86) #92

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6594195
doc: surface numerical ranking for 3D NS optimization
andrinr Jun 29, 2026
fef6564
doc: give optimization experiments readable section titles
andrinr Jul 9, 2026
4c1bab2
Merge branch 'main' into doc/ns3d-optimization-ranking
dionhaefner Jul 9, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 19 additions & 4 deletions docs/generate_results.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -195,6 +195,9 @@
"ic_recovery": "IC Recovery",
"param_recovery": "Parameter Recovery",
"topopt": "Topology Optimisation",
"drag_opt": "Drag Optimisation",
"conductivity_recovery_bfgs": "Conductivity Recovery (L-BFGS)",
"recovery_constant_ic_bfgs_proj": "Initial-Condition Recovery (L-BFGS)",
"recovery": "Recovery",
"recovery3d": "Recovery 3D",
"agreement3d": "Agreement 3D",
Expand Down Expand Up @@ -452,8 +455,18 @@ def _time(v: tuple[float, float] | None) -> str:
return header, [(s, _time(fv), _time(vjp.get(s))) for s, fv in rows]


# final-objective metric name by domain optimization experiment
_OPT_FINAL_KEYS = ("final_error", "final_drag", "final_compliance")
# final-objective metric name by domain optimization experiment, in priority
# order (the first key present in a result's metrics is used to rank it).
_OPT_FINAL_KEYS = ("final_error", "final_drag", "final_compliance", "final_ic_error")

# Human-readable column labels for the ranked metric. Falls back to a generic
# de-underscored title when a key is missing here.
_OPT_METRIC_LABELS = {
"final_error": "Final error",
"final_drag": "Final drag",
"final_compliance": "Final compliance",
"final_ic_error": "Final IC recovery error",
}


def _rank_optimization(problem: str) -> tuple[list[str], list[tuple]] | None:
Expand All @@ -472,14 +485,16 @@ def _rank_optimization(problem: str) -> tuple[list[str], list[tuple]] | None:
continue
val = m.get(key)
if isinstance(val, int | float) and math.isfinite(val):
metric_label = key.replace("final_", "final ").replace("_", " ")
metric_label = _OPT_METRIC_LABELS.get(
key, key.replace("final_", "final ").replace("_", " ").capitalize()
)
rows.append((r["solver"], float(val), bool(m.get("converged", False))))
if rows:
break
if len(rows) < 2:
return None
rows.sort(key=lambda t: t[1])
header = ["Solver", metric_label.capitalize(), "Converged"]
header = ["Solver", metric_label, "Converged"]
return header, [(s, _fmt_sci(v), "yes" if conv else "no") for s, v, conv in rows]


Expand Down
Loading