def test_all_benchmarks():
all_benchmarks(n_epochs=1, save_path='tests/data/')
def test_all_benchmarks(save_path):
all_benchmarks(n_epochs=1, save_path=save_path)
def test_all_benchmarks(save_path):
all_benchmarks(n_epochs=1, save_path=save_path, show_plot=False)
def test_all_benchmarks():
all_benchmarks(n_epochs=1)
def test_all_benchmarks():
all_benchmarks(n_epochs=1, unit_test=True)
help="whether to use cuda (will apply only if cuda is available")
parser.add_argument(
"--all",
action='store_true',
help="whether to use cuda (will apply only if cuda is available")
parser.add_argument(
"--benchmark",
action='store_true',
help="whether to use cuda (will apply only if cuda is available")
parser.add_argument("--url",
type=str,
help="the url for downloading gene_dataset")
args = parser.parse_args()
n_epochs = args.epochs
use_cuda = not args.nocuda
if args.all:
all_benchmarks(n_epochs=n_epochs, use_cuda=use_cuda)
elif args.harmonization:
harmonization_benchmarks(n_epochs=n_epochs, use_cuda=use_cuda)
elif args.annotation:
annotation_benchmarks(n_epochs=n_epochs, use_cuda=use_cuda)
else:
dataset = load_datasets(args.dataset, url=args.url)
model = available_models[args.model](
dataset.nb_genes, dataset.n_batches * args.nobatches,
dataset.n_labels)
inference_cls = VariationalInference if args.model == 'VAE' else JointSemiSupervisedVariationalInference
infer = inference_cls(model, dataset, use_cuda=use_cuda)
infer.train(n_epochs=n_epochs)
