algorithms/DANN.py

• 3 F841: local variable 'hparam_d_steps_per_g_step' is assigned to but never used (and 1 similar)

  • Local variable 'hparam_d_steps_per_g_step' is assigned to but never used:

    129 d_steps_per_g = hparam_d_steps_per_g_step = 1

  • Local variable 'num_classes' is assigned to but never used:

    191 num_classes = d_out

  • Local variable 'hparam_d_steps_per_g_step' is assigned to but never used:

    266 d_steps_per_g = hparam_d_steps_per_g_step = 1
• 15 E501: line too long (90 > 79 characters) (and 10 similar)

  • Line too long (90 > 79 characters):

    18 featurizer, classifier = initialize_model(config, d_out=d_out, is_featurizer=True)

  • Line too long (85 > 79 characters):

    32 assert config.num_train_domains <= 1000 # domain space shouldn't be too large

  • Line too long (83 > 79 characters):

    46 self.discriminator = torch.nn.Sequential(torch.nn.Linear(emb_dim, emb_dim),

  • Line too long (83 > 79 characters):

    48 torch.nn.Linear(emb_dim, emb_dim),

  • Line too long (118 > 79 characters):

    50 torch.nn.Linear(emb_dim, config.num_train_domains)).to(config.device)

  • Line too long (80 > 79 characters):

    117 disc_loss = F.cross_entropy(disc_out, disc_labels, reduction='none')

  • Line too long (89 > 79 characters):

    157 return self.loss.compute(results['y_pred'], results['y_true'], return_dict=False)

  • Line too long (115 > 79 characters):

    166 featurizer, pooler, classifier = initialize_model(config, d_out=d_out, is_featurizer=True, is_pooled=False)

  • Line too long (85 > 79 characters):

    181 assert config.num_train_domains <= 1000 # domain space shouldn't be too large

  • Line too long (99 > 79 characters):

    194 self.discriminator_gnn = GNN_node(num_layer=2, emb_dim=emb_dim, dropout=0, batchnorm=False,

  • Line too long (111 > 79 characters):

    195 dataset_group=config.model_kwargs['dataset_group']).to(config.device)

  • Line too long (101 > 79 characters):

    198 self.discriminator_mlp = torch.nn.Linear(emb_dim, config.num_train_domains).to(config.device)

  • Line too long (85 > 79 characters):

    201 # torch.nn.BatchNorm1d(emb_dim), torch.nn.ReLU(),

  • Line too long (88 > 79 characters):

    202 # torch.nn.Linear(emb_dim, config.num_train_domains)

  • Line too long (89 > 79 characters):

    294 return self.loss.compute(results['y_pred'], results['y_true'], return_dict=False)
• 5 E128: continuation line under-indented for visual indent
  • 52 self.featurizer.d_out).to(config.device)
  • 124 [disc_input], create_graph=True)[0]
  • 301 metric, n_train_steps, conditional=False, class_balance=False)
  • 309 metric, n_train_steps, conditional=True, class_balance=True)
  • 316 metric, n_train_steps)
• 4 E302: expected 2 blank lines, found 1
  • 11 class AbstractDANN(SingleModelAlgorithm):
  • 159 class OurAbstractDANN(SingleModelAlgorithm):
  • 296 class DANN(AbstractDANN):
  • 311 class DANNG(OurAbstractDANN):
• 3 E261: at least two spaces before inline comment
  • 32 assert config.num_train_domains <= 1000 # domain space shouldn't be too large
  • 170 model = classifier # fake, useless
  • 181 assert config.num_train_domains <= 1000 # domain space shouldn't be too large
• 2 E231: missing whitespace after ','
  • 112 disc_labels = move_to(metadata[:,0].flatten(), self.device)
  • 263 disc_labels = move_to(metadata[:,0].flatten(), self.device)
• 1 W292: no newline at end of file
  • 316 metric, n_train_steps)