aqora / Realization of a Quantum Streaming Algorithm on Long-lived Trapped-ion Qubits

Public

1

Dataset
Discussions

About dataset version

Streaming Hidden Matching Shot Data

Supplementary dataset for “Realization of a Quantum Streaming Algorithm on Long-lived Trapped-ion Qubits” (Niroula et al., Quantinuum & JPMorgan Chase, 2025). The study (arXiv:2511.03689) demonstrates a Hidden Matching (HM) streaming protocol on the Quantinuum H‑Series trapped-ion processor, showcasing an exponential memory advantage over classical sketches. This Parquet file aggregates every raw quantum shot used to produce the success/fidelity figures reported therein, so no additional metadata files are required.

Contents

  • Rows: 80 000 quantum shots
  • Structure:
    • 4 independent execution batches
    • 4 HM sizes (n = 4, 8, 16, 32)
    • 5 000 repetitions per size per batch

Experimental context

Every row captures a single execution of the streaming HM protocol:
  1. Vertices v and their binary labels x_v are streamed into a compact quantum sketch.
  2. Matching edges define the measurement basis (defining the hidden matching instance).
  3. A final projective measurement yields the sketch outcome (Π⁺, Π⁻, Π₀) recorded in measuredResults.
The dataset also tracks the measurement branch (meas) so fidelity calculations can focus on the meas = [1, 0] subset, and includes auxiliary observables (a, b) used internally for vertex and edge updates.

Schema

Below is the schema of the data. Each row represents one quantum shot.
ColumnTypeDescription
nIntegerHidden Matching problem size (4, 8, 16, or 32).
vertexListIntStreamed vertex sequence for the shot.
vertex_bitListIntBinary labels assigned to streamed vertices.
edge_1ListIntFirst endpoints of the streamed matching edges.
edge_2ListIntSecond endpoints of the streamed matching edges.
measuredResultsListIntProjective measurement result for the final sketch.
measListIntMeasurement-branch indicator ([1, 0] marks shots used for fidelity estimates).
aListIntAuxiliary observable for vertex-update tracking (nullable).
bListIntAuxiliary observable for edge-update tracking (nullable).
Arrays are stored as integer lists; when auxiliary data is absent the field is null.

Loading examples

Polars

import polars as pl
from aqora_cli.pyarrow import dataset

df = pl.scan_pyarrow_dataset(
    dataset("aqora/realization-of-a-quantum-streaming-algorithm-on-long-lived-trapped-ion-qubits", "v0.0.0")
).collect()

stats = (
    df.groupby("n")
      .agg([
          pl.len().alias("shots"),
          pl.col("measuredResults").list.last().mean().alias("avg_last_bit"),
      ])
      .sort("n")
)
print(stats)

Pandas

import pandas as pd

df = pd.read_parquet("aqora://aqora/realization-of-a-quantum-streaming-algorithm-on-long-lived-trapped-ion-qubits/v0.0.0")
print(df.head())

Suggested analyses

  • Reproduce success/failure ratios by filtering meas == [1, 0] and comparing measuredResults against the HM ground truth.
  • Inspect vertex or edge streams for specific n to study how classical sketches diverge from the experiment.
  • Explore correlations between auxiliary observables (a, b) and sketch fidelity to identify hardware drift.
For full experimental details—circuit construction, streaming protocol, and theoretical guarantees—consult arXiv:2511.03689. This README serves as the Aqora dataset card for the shot-level Parquet artifact.