Skip to content

Decode detector samples with BP-OSD

Install the default decoder interfaces:

Terminal window
pip install bloqade-decoders

Describe the errors

Section titled “Describe the errors”

A Stim detector error model gives each possible error a probability and the detector or logical-observable bits it flips. This small model has two detectors and one logical observable:

import stim


dem = stim.DetectorErrorModel("""
    error(0.1) D0
    error(0.1) D0 D1
    error(0.1) D1 L0
""")

Decode a batch

Section titled “Decode a batch”

Pass the model to BpOsdDecoder, then provide one row per shot. The return value contains predicted observable corrections in the same row order:

import numpy as np
from bloqade.decoders import BpOsdDecoder


decoder = BpOsdDecoder(dem)
syndromes = np.array([[False, False], [False, True]], dtype=bool)
corrections = decoder.decode(syndromes)
print(corrections)

The same batch shape works with the other decoder interfaces. See the upstream Quick Start for lookup-table and optional Gurobi decoders, or the Overview for package extras and usage notes.