def __init__(self, population, sample_size, one_step_proportion=1.0):
self.one_step_proportion = one_step_proportion
self.sample_size = sample_size
self.population = population
self.neighbours = Collection(population.factory)
lmbda = -mathln(self.one_step_proportion)
# We simple iterate over all networks and add to our neighbourhood
for n in population:
if one_step_proportion < 1.0:
mutes = np.random.poisson(lmbda, self.sample_size) + 1
else:
mutes = np.ones(self.sample_size, dtype=int)
network_neighbours = Collection(population.factory)
network_neighbours.fill_with_mutations(n, mutes)
# We just add to the main population
self.neighbours.extend(network_neighbours)
def __init__(self, center, sample_size, one_step_proportion=1.0):
# Sample from the neighbourhood
#
# @center: our focal network.
# @one_step_proportion: the proportion of the sample that we want to
# have only one step mutations
#
self.one_step_proportion = one_step_proportion
self.sample_size = sample_size
self.center = center
if one_step_proportion < 1.0:
# Set lambda so we get the right proportion of one-step neighbours
lmbda = -mathln(self.one_step_proportion)
# We want at least one mutation, so add one. Note that the proportion
# of these that equals 1 should be one_step_proportion.
self.mutations = np.random.poisson(lmbda, self.sample_size) + 1
else:
self.mutations = np.ones(self.sample_size, dtype=int)
self.neighbours = Collection(center.factory)
self.neighbours.fill_with_mutations(center, self.mutations)