def get_dataloader(
split: str = 'train',
dataset_args: dict = {},
dataloader_args: dict = {},
weighted_sampling: Optional[bool] = False
) -> torch.utils.data.DataLoader:
"""Loads dataloaders for the PE CT dataset.
Args:
split (str, optional): the split with which to filter the CSV.
dataset_args (dict, optional): keyword arguments for dataset init.
dataloader_args (dict): keyword arguments for dataloader init.
weighted_sampling (bool, optional):
Returns:
a pytorch dataloader
"""
dataset = RSNADataset(split=split, **dataset_args)
# weighted sampling
if weighted_sampling and split == 'train':
sampler = utils.weighted_sampler(dataset)
dataloader_args['shuffle'] = False
dataloader = DataLoader(dataset, sampler=sampler, **dataloader_args)
else:
dataloader = DataLoader(dataset, **dataloader_args)
return dataloader
def genetic_algorithm(population,
fitness_fn,
gene_pool=[0, 1],
f_thres=None,
ngen=1000,
pmut=0.1): # noqa
"""[Figure 4.8]"""
for i in range(ngen):
new_population = []
fitnesses = map(fitness_fn, population)
random_selection = weighted_sampler(population, fitnesses)
for j in range(len(population)):
x = random_selection()
y = random_selection()
child = reproduce(x, y)
if random.uniform(0, 1) < pmut:
child = mutate(child, gene_pool)
new_population.append(child)
population = new_population
if f_thres:
fittest_individual = argmax(population, key=fitness_fn)
if fitness_fn(fittest_individual) >= f_thres:
return fittest_individual
return argmax(population, key=fitness_fn)
def sample(self):
"""Return a random sample from the distribution."""
if self.sampler is None:
self.sampler = weighted_sampler(list(self.dictionary.keys()),
list(self.dictionary.values()))
return self.sampler()
def selection_chances(fitness_fn, population):
fitnesses = map(fitness_fn, population)
return weighted_sampler(population, fitnesses)
def sample(self):
"Return a random sample from the distribution."
if self.sampler is None:
self.sampler = weighted_sampler(list(self.dictionary.keys()),
list(self.dictionary.values()))
return self.sampler()
def select(r, population, fitness_fn):
fitnesses = map(fitness_fn, population)
sampler = weighted_sampler(population, fitnesses)
return [sampler() for i in range(r)]
