def candidate_generator(self, seed_graphs):
"""Generate candidates.
Parameters
----------
seed_graphs : networkx graphs
The iterator over the seed graphs, i.e. the graphs that are used as
a starting point for the proposal.
"""
start = time.time()
graphs = transform(seed_graphs,
program=AnnotateImportance(
program=self.fit_wrapped_predictor.program))
graphs = list(graphs)
logger.debug('Working on %d graphs' % len(graphs))
# mark the position of nodes with the attribute 'exclude' to remove
# the influence of primers
graphs = transform(
graphs,
program=MarkWithIntervals(quadruples=self.exclusion_quadruples))
# find the ktop largest (reverse=True) values for the
# attribute='importance' in the vertices of a graph
# and add an attribute to each vertex that is 'selected'=True
# if the node is among the ktop
graphs = transform(graphs,
program=MarkKTop(attribute='importance',
exclude_attribute='exclude',
ktop=self.n_substitutions,
reverse=True,
mark_attribute='selected'))
# generate graphs that have all possible combination of symbols in
# the nodes marked by MarkTop
graphs = transform(graphs,
program=ReplaceWithAllCombinations(
attribute='selected',
label_list=self.label_list))
# refold the sequences to account for structural changes
graphs = transform(graphs, program=self.seq_to_structure_prog)
# return the candidate graphs
candidate_graphs = list(graphs)
delta_time = datetime.timedelta(seconds=(time.time() - start))
logger.info('Candidate generation took: %s' % (str(delta_time)))
logger.info('Number of candidates: %d' % (len(candidate_graphs)))
return candidate_graphs
def candidate_generator(self, seed_graphs):
"""Generate candidates.
Parameters
----------
seed_graphs : networkx graphs
The iterator over the seed graphs, i.e. the graphs that are used as
a starting point for the proposal.
"""
start = time.time()
graphs = transform(seed_graphs,
program=AnnotateImportance(
program=self.fit_wrapped_predictor.program))
graphs = list(graphs)
logger.debug('Working on %d graphs' % len(graphs))
# mark the position of nodes with the attribute 'exclude' to remove
# the influence of primers
graphs = transform(graphs,
program=MarkWithIntervals(
quadruples=self.exclusion_quadruples))
# find the ktop largest (reverse=True) values for the
# attribute='importance' in the vertices of a graph
# and add an attribute to each vertex that is 'selected'=True
# if the node is among the ktop
graphs = transform(graphs,
program=MarkKTop(attribute='importance',
exclude_attribute='exclude',
ktop=self.n_substitutions,
reverse=True,
mark_attribute='selected'))
# generate graphs that have all possible combination of symbols in
# the nodes marked by MarkTop
graphs = transform(graphs, program=ReplaceWithAllCombinations(
attribute='selected', label_list=self.label_list))
# refold the sequences to account for structural changes
graphs = transform(graphs, program=self.seq_to_structure_prog)
# return the candidate graphs
candidate_graphs = list(graphs)
delta_time = datetime.timedelta(seconds=(time.time() - start))
logger.info('Candidate generation took: %s' % (str(delta_time)))
logger.info('Number of candidates: %d' % (len(candidate_graphs)))
return candidate_graphs
def transform(self, orig_graphs=None):
"""transform."""
try:
graphs = self._transform(orig_graphs)
# reduce all 'label' attributes of contracted nodes to a
# histogram to be written in the 'label' attribute of the
# resulting graph
label_modifier = contraction_modifier(attribute_in='label',
attribute_out='label',
reduction='categorical')
# reduce all 'weight' attributes of contracted nodes using
# a sum to be written in the 'weight' attribute of the
# resulting graph
weight_modifier = contraction_modifier(attribute_in='weight',
attribute_out='weight',
reduction='sum')
modifiers = [label_modifier, weight_modifier]
s = self.original_edges_to_nesting
priors = dict(nesting=self.nesting,
weight_scaling_factor=1,
original_edges_to_nesting=s)
ca = 'max_clique_hash'
graphs = transform(graphs,
program=Contract(modifiers=modifiers,
contraction_attribute=ca),
parameters_priors=priors)
return graphs
except Exception as e:
logger.debug('Failed iteration. Reason: %s' % e)
logger.debug('Exception', exc_info=True)
def transform(self, orig_graphs=None):
"""transform."""
try:
graphs = self._transform(orig_graphs)
# reduce all 'label' attributes of contracted nodes to a
# histogram to be written in the 'label' attribute of the
# resulting graph
label_modifier = contraction_modifier(attribute_in='label',
attribute_out='label',
reduction='categorical')
# reduce all 'weight' attributes of contracted nodes using
# a sum to be written in the 'weight' attribute of the
# resulting graph
weight_modifier = contraction_modifier(attribute_in='weight',
attribute_out='weight',
reduction='sum')
modifiers = [label_modifier, weight_modifier]
s = self.original_edges_to_nesting
priors = dict(nesting=self.nesting,
weight_scaling_factor=1,
original_edges_to_nesting=s)
ca = 'max_clique_hash'
graphs = transform(graphs,
program=Contract(modifiers=modifiers,
contraction_attribute=ca),
parameters_priors=priors)
return graphs
except Exception as e:
logger.debug('Failed iteration. Reason: %s' % e)
logger.debug('Exception', exc_info=True)
def efficient_selection(self,
candidate_graphs,
known_graphs=None):
"""Propose a small number of alternative structures.
Parameters
----------
candidate_graphs : networkx graphs
The iterator over the seed graphs, i.e. the graphs that are used
as a starting point for the proposal.
known_graphs : networkx graphs
The iterator over the already known graphs. These are used to bias
the exploration towards less similar proposals.
"""
start = time.time()
candidate_graphs = transform(
candidate_graphs,
program=AnnotateImportance(
program=self.fit_wrapped_predictor.program))
candidate_graphs = list(candidate_graphs)
# transform graphs according to importance
# this allows similarity notion to be task dependent
known_graphs = transform(
known_graphs,
program=AnnotateImportance(
program=self.fit_wrapped_predictor.program))
known_graphs = list(known_graphs)
# store the nearest neighbors in knn_manager
# compute the k nearest neighbors distances of each proposal graph
knn_manager = KNNManager(n_neighbors=self.n_neighbors, complexity=3)
knn_manager.setup(known_graphs=known_graphs,
candidate_graphs=candidate_graphs)
delta_time = datetime.timedelta(seconds=(time.time() - start))
logger.info('Knn computation took: %s' % (str(delta_time)))
# compute predictions
predicted_graphs = predict(candidate_graphs,
program=self.fit_wrapped_predictor)
predicted_graphs = list(predicted_graphs)
scores = np.array([graph.graph['score']
for graph in predicted_graphs]).reshape(-1, 1)
# iterations
tradeoff = self.exploration_vs_exploitation_tradeoff
selection_ids = []
for i in range(self.n_proposals):
uncertainties = knn_manager.average_distances()
# run the acquisition function (n_proposals times)
# and return best_id
maximal_id = self._acquisition(
scores,
uncertainties,
exploration_vs_exploitation_tradeoff=tradeoff)
# update distances with new selection
knn_manager.add_element(maximal_id)
# store id
selection_ids.append(maximal_id)
graph = candidate_graphs[maximal_id]
logger.debug('>%s' % graph.graph['header'])
logger.debug(graph.graph['sequence'])
return selection_ids
def efficient_selection(self,
candidate_graphs,
known_graphs=None):
"""Propose a small number of alternative structures.
Parameters
----------
candidate_graphs : networkx graphs
The iterator over the seed graphs, i.e. the graphs that are used
as a starting point for the proposal.
known_graphs : networkx graphs
The iterator over the already known graphs. These are used to bias
the exploration towards less similar proposals.
"""
start = time.time()
candidate_graphs = transform(
candidate_graphs,
program=AnnotateImportance(
program=self.fit_wrapped_predictor.program))
candidate_graphs = list(candidate_graphs)
# transform graphs according to importance
# this allows similarity notion to be task dependent
known_graphs = transform(
known_graphs,
program=AnnotateImportance(
program=self.fit_wrapped_predictor.program))
known_graphs = list(known_graphs)
# store the nearest neighbors in knn_manager
# compute the k nearest neighbors distances of each proposal graph
knn_manager = KNNManager(n_neighbors=self.n_neighbors, complexity=3)
knn_manager.setup(known_graphs=known_graphs,
candidate_graphs=candidate_graphs)
delta_time = datetime.timedelta(seconds=(time.time() - start))
logger.info('Knn computation took: %s' % (str(delta_time)))
# compute predictions
predicted_graphs = predict(candidate_graphs,
program=self.fit_wrapped_predictor)
predicted_graphs = list(predicted_graphs)
scores = np.array([graph.graph['score']
for graph in predicted_graphs]).reshape(-1, 1)
# iterations
tradeoff = self.exploration_vs_exploitation_tradeoff
selection_ids = []
for i in range(self.n_proposals):
uncertainties = knn_manager.average_distances()
# run the acquisition function (n_proposals times)
# and return best_id
maximal_id = self._acquisition(
scores,
uncertainties,
exploration_vs_exploitation_tradeoff=tradeoff)
# update distances with new selection
knn_manager.add_element(maximal_id)
# store id
selection_ids.append(maximal_id)
graph = candidate_graphs[maximal_id]
logger.debug('>%s' % graph.graph['header'])
logger.debug(graph.graph['sequence'])
return selection_ids