Results and plotting helpers

Quon’s compiler emits machine-readable JSON (counts from the Aer bridge, --emit-resource-report, --emit-na-schedule), but reading raw JSON by eye is slow. python/quon_viz.py (issue #196) is a small, dependency-light presentation layer that turns those artifacts into bar charts and readable tables. It mirrors the names you already know from Qiskit so a migration is a one-line swap.

Install the deps once (matplotlib is the only new one):

pip install -r python/requirements.txt

Instead of qiskit.visualization.plot_histogram

import sys
sys.path.insert(0, "python")
import quon_viz

counts = {"00": 2050, "11": 2022}
quon_viz.plot_histogram(counts, title="Bell state", bar_labels=True,
                        filename="bell.png")

plot_histogram accepts the same keyword shape as Qiskit’s — figsize, color, title, legend_keys, bar_labels, number_to_keep, sort — but is safe to call headless: it forces the non-interactive Agg backend and never calls plt.show(). Pass filename= to save a PNG and return None; omit it to get the Figure back for further tweaking. The verifier scripts test/verify/bell.py and test/verify/grover.py already use it — each writes a histogram under $QUON_VIZ_DIR (or $TMPDIR/quon_viz) when run.

Pretty-print compiler metrics

metrics_table formats any flat (or lightly nested) metrics dict as an aligned two-column table — useful for a quick depth / gate-count / SWAP readout from your own benchmark loop:

print(quon_viz.metrics_table({"depth": 12, "swaps": 3, "cx": 8,
                               "single_qubit_gates": 7, "fidelity": 0.9895}))

Neutral-atom resource reports

quonc --emit-resource-report emits a flat JSON object. summarize_na_report groups it into Resource / Qubits / Single-qubit gates / Fidelity blocks and appends the error_budget and temporal_atom_metrics sub-objects:

import json, quon_viz
report = json.load(open("report.json"))
print(quon_viz.summarize_na_report(report))

It accepts a parsed dict, a path to a JSON file, or a JSON string, so the shortest form is just quon_viz.summarize_na_report("report.json").

Neutral-atom schedule timelines

summarize_na_schedule reads the na_schedule_view envelope from --emit-na-schedule and prints one compact line per cycle (Move(2), Entangle2(a0,a1), Transfer(a0,SlmToAod), rz(a1), …), with a header summarizing zones and headline metrics. max_layers= truncates the per-cycle listing while keeping the full-schedule metrics block:

print(quon_viz.summarize_na_schedule("schedule.json", max_layers=10))

End-to-end NA summary

samples/research/na_resource_summary.py compiles a .qn program with quonc and prints both summaries in one shot — a runnable readout of what a neutral-atom compile produced:

QUONC=target/release/quonc python samples/research/na_resource_summary.py \
    test/na/qaoa_graph.qn

Optional: Bloch sphere

plot_bloch(statevector, ...) thin-wraps qiskit.visualization.plot_bloch_multivector for the headless case (forces Agg, saves to filename=). It is optional: if qiskit.visualization is unavailable it prints a notice and returns None rather than failing.

quon_viz.plot_bloch([1, 0], title="|0>", filename="bloch.png")

CLI shortcut

The module is also runnable as a pretty-printer for a report or schedule file:

python python/quon_viz.py report path/to/report.json
python python/quon_viz.py schedule path/to/schedule.json --max-layers 8