def test_trainer_struct(self):
output_dir = 'output'
args = [
'--train-pairs',
get_data_path('train.txt'), '--test-pairs',
get_data_path('test.txt'), '--valid-pairs',
get_data_path('valid.txt'), '--output-directory', output_dir,
'--epochs', '1', '--batch-size', '1', '--num-workers', '1',
'--learning-rate', '1e-4', '--clip-ends', 'False',
'--visualization-fraction', '1', '--gpus', '1'
]
parser = argparse.ArgumentParser(add_help=False)
parser = LightningAligner.add_model_specific_args(parser)
parser.add_argument('--num-workers', type=int)
parser.add_argument('--gpus', type=int)
args = parser.parse_args(args)
model = LightningAligner(args)
trainer = Trainer(
max_epochs=args.epochs,
gpus=args.gpus,
check_val_every_n_epoch=1,
# profiler=profiler,
fast_dev_run=True,
# auto_scale_batch_size='power'
)
trainer.fit(model)
def test_decoding2(self):
X = 'HECDRKTCDESFSTKGNLRVHKLGH'
Y = 'LKCSGCGKNFKSQYAYKRHEQTH'
needle = NeedlemanWunschDecoder(self.operator)
dm = torch.Tensor(np.loadtxt(get_data_path('dm.txt')))
decoded = needle.traceback(dm)
pred_x, pred_y, pred_states = list(zip(*decoded))
states2alignment(np.array(pred_states), X, Y)
def setUp(self):
hmm = get_data_path('ABC_tran.hmm')
n_alignments = 100
align_df = hmm_alignments(n=40,
seed=0,
n_alignments=n_alignments,
hmmfile=hmm)
cols = [
'chain1_name', 'chain2_name', 'tmscore1', 'tmscore2', 'rmsd',
'chain1', 'chain2', 'alignment'
]
align_df.columns = cols
parts = n_alignments // 10
train_df = align_df.iloc[:parts * 8]
test_df = align_df.iloc[parts * 8:parts * 9]
valid_df = align_df.iloc[parts * 9:]
# save the files to disk.
train_df.to_csv('train.txt', sep='\t', index=None, header=None)
test_df.to_csv('test.txt', sep='\t', index=None, header=None)
valid_df.to_csv('valid.txt', sep='\t', index=None, header=None)
def setUp(self):
self.data_path = get_data_path('test_tm_align.tab')
self.tokenizer = UniprotTokenizer(pad_ends=False)
def setUp(self):
self.data_path = get_data_path('example.txt')
self.pairs = pd.read_table(self.data_path, header=None)