def test_sampleset1_append1(self):
sampleset = dimod.SampleSet.from_samples({'a': -1, 'b': 1}, dimod.SPIN, energy=0)
new = dimod.append_variables(sampleset, {'c': -1, 'd': -1})
target = dimod.SampleSet.from_samples({'a': -1, 'b': 1, 'c': -1, 'd': -1}, dimod.SPIN, energy=0)
self.assertEqual(new, target)
def test_two_samplesets(self):
sampleset0 = dimod.SampleSet.from_samples([{'a': -1, 'b': 1}, {'a': -1, 'b': -1}],
dimod.SPIN, energy=[-2, 2])
sampleset1 = dimod.SampleSet.from_samples([{'c': -1, 'd': 1}, {'c': -1, 'd': -1}],
dimod.SPIN, energy=[-1, 1])
target = dimod.SampleSet.from_samples([{'a': -1, 'b': 1, 'c': -1, 'd': 1},
{'a': -1, 'b': -1, 'c': -1, 'd': -1}],
dimod.SPIN, energy=[-2, 2])
self.assertEqual(dimod.append_variables(sampleset0, sampleset1), target)
def test_overlapping_variables(self):
sampleset = dimod.SampleSet.from_samples([{'a': -1, 'b': 1}, {'a': -1, 'b': -1}],
dimod.SPIN, energy=0)
with self.assertRaises(ValueError):
dimod.append_variables(sampleset, [{'c': -1, 'd': -1, 'a': -1}])
def sample_markov_network(MN,
sampler=None,
fixed_variables=None,
return_sampleset=False,
**sampler_args):
"""Samples from a markov network using the provided sampler.
Parameters
----------
G : NetworkX graph
A Markov Network as returned by :func:`.markov_network`
sampler
A binary quadratic model sampler. A sampler is a process that
samples from low energy states in models defined by an Ising
equation or a Quadratic Unconstrained Binary Optimization
Problem (QUBO). A sampler is expected to have a 'sample_qubo'
and 'sample_ising' method. A sampler is expected to return an
iterable of samples, in order of increasing energy. If no
sampler is provided, one must be provided using the
`set_default_sampler` function.
fixed_variables : dict
A dictionary of variable assignments to be fixed in the markov network.
return_sampleset : bool (optional, default=False)
If True, returns a :obj:`dimod.SampleSet` rather than a list of samples.
**sampler_args
Additional keyword parameters are passed to the sampler.
Returns
-------
samples : list[dict]/:obj:`dimod.SampleSet`
A list of samples ordered from low-to-high energy or a sample set.
Examples
--------
>>> import dimod
...
>>> potentials = {('a', 'b'): {(0, 0): -1,
... (0, 1): .5,
... (1, 0): .5,
... (1, 1): 2}}
>>> MN = dnx.markov_network(potentials)
>>> sampler = dimod.ExactSolver()
>>> samples = dnx.sample_markov_network(MN, sampler)
>>> samples[0] # doctest: +SKIP
{'a': 0, 'b': 0}
>>> import dimod
...
>>> potentials = {('a', 'b'): {(0, 0): -1,
... (0, 1): .5,
... (1, 0): .5,
... (1, 1): 2}}
>>> MN = dnx.markov_network(potentials)
>>> sampler = dimod.ExactSolver()
>>> samples = dnx.sample_markov_network(MN, sampler, return_sampleset=True)
>>> samples.first # doctest: +SKIP
Sample(sample={'a': 0, 'b': 0}, energy=-1.0, num_occurrences=1)
>>> import dimod
...
>>> potentials = {('a', 'b'): {(0, 0): -1,
... (0, 1): .5,
... (1, 0): .5,
... (1, 1): 2},
... ('b', 'c'): {(0, 0): -9,
... (0, 1): 1.2,
... (1, 0): 7.2,
... (1, 1): 5}}
>>> MN = dnx.markov_network(potentials)
>>> sampler = dimod.ExactSolver()
>>> samples = dnx.sample_markov_network(MN, sampler, fixed_variables={'b': 0})
>>> samples[0] # doctest: +SKIP
{'a': 0, 'c': 0, 'b': 0}
Notes
-----
Samplers by their nature may not return the optimal solution. This
function does not attempt to confirm the quality of the returned
sample.
"""
bqm = markov_network_bqm(MN)
if fixed_variables:
# we can modify in-place since we just made it
bqm.fix_variables(fixed_variables)
sampleset = sampler.sample(bqm, **sampler_args)
if fixed_variables:
# add the variables back in
sampleset = dimod.append_variables(sampleset, fixed_variables)
if return_sampleset:
return sampleset
else:
return list(map(dict, sampleset.samples()))