def chain_break_frequency(samples_like, embedding):
"""Determine the frequency of chain breaks in the given samples.
Args:
samples_like (samples_like/:obj:`dimod.SampleSet`):
A collection of raw samples. 'samples_like' is an extension of NumPy's array_like.
See :func:`dimod.as_samples`.
embedding (dict):
Mapping from source graph to target graph as a dict of form {s: {t, ...}, ...},
where s is a source-model variable and t is a target-model variable.
Returns:
dict: Frequency of chain breaks as a dict in the form {s: f, ...}, where s
is a variable in the source graph and float f the fraction
of broken chains.
Examples:
This example embeds a single source node, 'a', as a chain of two target nodes (0, 1)
and uses :func:`.chain_break_frequency` to show that out of two synthetic samples,
one ([-1, +1]) represents a broken chain.
>>> import numpy as np
...
>>> samples = np.array([[-1, +1], [+1, +1]])
>>> embedding = {'a': {0, 1}}
>>> print(dwave.embedding.chain_break_frequency(samples, embedding)['a'])
0.5
"""
if isinstance(samples_like, dimod.SampleSet):
labels = samples_like.variables
samples = samples_like.record.sample
num_occurrences = samples_like.record.num_occurrences
else:
samples, labels = dimod.as_samples(samples_like)
num_occurrences = np.ones(samples.shape[0])
if not all(v == idx for idx, v in enumerate(labels)):
labels_to_idx = {v: idx for idx, v in enumerate(labels)}
embedding = {
v: {labels_to_idx[u]
for u in chain}
for v, chain in embedding.items()
}
if not embedding:
return {}
variables, chains = zip(*embedding.items())
broken = broken_chains(samples, chains)
return {
v: float(np.average(broken[:, cidx], weights=num_occurrences))
for cidx, v in enumerate(variables)
}
def chain_break_frequency(samples_like, embedding):
"""Determine the frequency of chain breaks in the given samples.
Args:
samples_like (samples_like/:obj:`dimod.SampleSet`):
A collection of raw samples. 'samples_like' is an extension of NumPy's array_like.
See :func:`dimod.as_samples`.
embedding (dict):
Mapping from source graph to target graph as a dict of form {s: {t, ...}, ...},
where s is a source-model variable and t is a target-model variable.
Returns:
dict: Frequency of chain breaks as a dict in the form {s: f, ...}, where s
is a variable in the source graph, and frequency, a float, is the fraction
of broken chains.
Examples:
This example embeds a single source node, 'a', as a chain of two target nodes (0, 1)
and uses :func:`.chain_break_frequency` to show that out of two synthetic samples,
one ([-1, +1]) represents a broken chain.
>>> import dimod
>>> import numpy as np
>>> samples = np.array([[-1, +1], [+1, +1]])
>>> embedding = {'a': {0, 1}}
>>> print(dimod.chain_break_frequency(samples, embedding)['a'])
0.5
This example embeds a single source node (0) as a chain of two target nodes (a, b)
and uses :func:`.chain_break_frequency` to show that out of two samples in a
dimod response, one ({'a': 1, 'b': 0}) represents a broken chain.
>>> import dimod
...
>>> response = dimod.SampleSet.from_samples([{'a': 1, 'b': 0}, {'a': 0, 'b': 0}],
... {'energy': [1, 0]}, {}, dimod.BINARY)
>>> embedding = {0: {'a', 'b'}}
>>> print(dimod.chain_break_frequency(response, embedding)[0])
0.5
"""
if isinstance(samples_like, dimod.SampleSet):
labels = samples_like.variables
samples = samples_like.record.sample
else:
samples, labels = dimod.as_samples(samples_like)
if not all(v == idx for idx, v in enumerate(labels)):
labels_to_idx = {v: idx for idx, v in enumerate(labels)}
embedding = {
v: {labels_to_idx[u]
for u in chain}
for v, chain in embedding.items()
}
if not embedding:
return {}
variables, chains = zip(*embedding.items())
broken = broken_chains(samples, chains)
freq = {
v: float(broken[:, cidx].mean())
for cidx, v in enumerate(variables)
}
return freq
def unembed_sampleset(target_sampleset,
embedding,
source_bqm,
chain_break_method=None,
chain_break_fraction=False,
return_embedding=False):
"""Unembed a sample set.
Given samples from a target binary quadratic model (BQM), construct a sample
set for a source BQM by unembedding.
Args:
target_sampleset (:obj:`dimod.SampleSet`):
Sample set from the target BQM.
embedding (dict):
Mapping from source graph to target graph as a dict of form
{s: {t, ...}, ...}, where s is a source variable and t is a target
variable.
source_bqm (:obj:`~dimod.BinaryQuadraticModel`):
Source BQM.
chain_break_method (function/list, optional):
Method or methods used to resolve chain breaks. If multiple methods
are given, the results are concatenated and a new field called
"chain_break_method" specifying the index of the method is appended
to the sample set.
Defaults to :func:`~dwave.embedding.chain_breaks.majority_vote`.
See :mod:`dwave.embedding.chain_breaks`.
chain_break_fraction (bool, optional, default=False):
Add a `chain_break_fraction` field to the unembedded :obj:`dimod.SampleSet`
with the fraction of chains broken before unembedding.
return_embedding (bool, optional, default=False):
If True, the embedding is added to :attr:`dimod.SampleSet.info`
of the returned sample set. Note that if an `embedding` key
already exists in the sample set then it is overwritten.
Returns:
:obj:`.SampleSet`: Sample set in the source BQM.
Examples:
This example unembeds from a square target graph samples of a triangular
source BQM.
>>> # Triangular binary quadratic model and an embedding
>>> J = {('a', 'b'): -1, ('b', 'c'): -1, ('a', 'c'): -1}
>>> bqm = dimod.BinaryQuadraticModel.from_ising({}, J)
>>> embedding = {'a': [0, 1], 'b': [2], 'c': [3]}
>>> # Samples from the embedded binary quadratic model
>>> samples = [{0: -1, 1: -1, 2: -1, 3: -1}, # [0, 1] is unbroken
... {0: -1, 1: +1, 2: +1, 3: +1}] # [0, 1] is broken
>>> energies = [-3, 1]
>>> embedded = dimod.SampleSet.from_samples(samples, dimod.SPIN, energies)
>>> # Unembed
>>> samples = dwave.embedding.unembed_sampleset(embedded, embedding, bqm)
>>> samples.record.sample # doctest: +SKIP
array([[-1, -1, -1],
[ 1, 1, 1]], dtype=int8)
"""
if chain_break_method is None:
chain_break_method = majority_vote
elif isinstance(chain_break_method, abc.Sequence):
# we want to apply multiple CBM and then combine
samplesets = [
unembed_sampleset(target_sampleset,
embedding,
source_bqm,
chain_break_method=cbm,
chain_break_fraction=chain_break_fraction)
for cbm in chain_break_method
]
sampleset = dimod.sampleset.concatenate(samplesets)
# Add a new data field tracking which came from
# todo: add this functionality to dimod
cbm_idxs = np.empty(len(sampleset), dtype=int)
start = 0
for i, ss in enumerate(samplesets):
cbm_idxs[start:start + len(ss)] = i
start += len(ss)
new = np.lib.recfunctions.append_fields(sampleset.record,
'chain_break_method',
cbm_idxs,
asrecarray=True,
usemask=False)
return type(sampleset)(new, sampleset.variables, sampleset.info,
sampleset.vartype)
variables = list(source_bqm.variables) # need this ordered
try:
chains = [embedding[v] for v in variables]
except KeyError:
raise ValueError("given bqm does not match the embedding")
record = target_sampleset.record
unembedded, idxs = chain_break_method(target_sampleset, chains)
reserved = {'sample', 'energy'}
vectors = {
name: record[name][idxs]
for name in record.dtype.names if name not in reserved
}
if chain_break_fraction:
broken = broken_chains(target_sampleset, chains)
if broken.size:
vectors['chain_break_fraction'] = broken.mean(axis=1)[idxs]
else:
vectors['chain_break_fraction'] = 0
info = target_sampleset.info.copy()
if return_embedding:
embedding_context = dict(
embedding=embedding,
chain_break_method=chain_break_method.__name__)
info.update(embedding_context=embedding_context)
return dimod.SampleSet.from_samples_bqm((unembedded, variables),
source_bqm,
info=info,
**vectors)
def unembed_sampleset(target_sampleset,
embedding,
source_bqm,
chain_break_method=None,
chain_break_fraction=False,
return_embedding=False):
"""Unembed a samples set.
Given samples from a target binary quadratic model (BQM), construct a sample
set for a source BQM by unembedding.
Args:
target_sampleset (:obj:`dimod.SampleSet`):
Sample set from the target BQM.
embedding (dict):
Mapping from source graph to target graph as a dict of form
{s: {t, ...}, ...}, where s is a source variable and t is a target
variable.
source_bqm (:obj:`dimod.BinaryQuadraticModel`):
Source BQM.
chain_break_method (function, optional):
Method used to resolve chain breaks.
See :mod:`dwave.embedding.chain_breaks`.
chain_break_fraction (bool, optional, default=False):
Add a `chain_break_fraction` field to the unembedded :obj:`dimod.SampleSet`
with the fraction of chains broken before unembedding.
return_embedding (bool, optional, default=False):
If True, the embedding is added to :attr:`dimod.SampleSet.info`
of the returned sample set. Note that if an `embedding` key
already exists in the sample set then it is overwritten.
Returns:
:obj:`.SampleSet`: Sample set in the source BQM.
Examples:
This example unembeds from a square target graph samples of a triangular
source BQM.
>>> # Triangular binary quadratic model and an embedding
>>> J = {('a', 'b'): -1, ('b', 'c'): -1, ('a', 'c'): -1}
>>> bqm = dimod.BinaryQuadraticModel.from_ising({}, J)
>>> embedding = {'a': [0, 1], 'b': [2], 'c': [3]}
>>> # Samples from the embedded binary quadratic model
>>> samples = [{0: -1, 1: -1, 2: -1, 3: -1}, # [0, 1] is unbroken
... {0: -1, 1: +1, 2: +1, 3: +1}] # [0, 1] is broken
>>> energies = [-3, 1]
>>> embedded = dimod.SampleSet.from_samples(samples, dimod.SPIN, energies)
>>> # Unembed
>>> samples = dwave.embedding.unembed_sampleset(embedded, embedding, bqm)
>>> samples.record.sample # doctest: +SKIP
array([[-1, -1, -1],
[ 1, 1, 1]], dtype=int8)
"""
if chain_break_method is None:
chain_break_method = majority_vote
variables = list(source_bqm.variables) # need this ordered
try:
chains = [embedding[v] for v in variables]
except KeyError:
raise ValueError("given bqm does not match the embedding")
record = target_sampleset.record
unembedded, idxs = chain_break_method(target_sampleset, chains)
reserved = {'sample', 'energy'}
vectors = {
name: record[name][idxs]
for name in record.dtype.names if name not in reserved
}
if chain_break_fraction:
vectors['chain_break_fraction'] = broken_chains(
target_sampleset, chains).mean(axis=1)[idxs]
info = target_sampleset.info.copy()
if return_embedding:
embedding_context = dict(
embedding=embedding,
chain_break_method=chain_break_method.__name__)
info.update(embedding_context=embedding_context)
return dimod.SampleSet.from_samples_bqm((unembedded, variables),
source_bqm,
info=info,
**vectors)