def test_sign_cora():
args = get_default_args()
args.task = "node_classification"
args.model = "sign"
args.dataset = "cora"
args.lr = 0.00005
args.hidden_size = 2048
args.num_layers = 3
args.num_propagations = 3
args.dropout = 0.3
args.directed = False
args.dropedge_rate = 0.2
args.asymm_norm = False
args.set_diag = False
args.remove_diag = False
task = build_task(args)
ret = task.train()
assert 0 < ret["Acc"] < 1
def objective(self, trials):
fixed_params = self.func_fixed()
params = self.func_search(trials)
params.update(fixed_params)
args = self.build_args(params)
model = build_model(args)
task = build_task(args, model=model, dataset=self.dataset)
result = task.train()
for item in list(result.keys()):
if "Val" in item:
result = {item: result[item]}
break
assert type(result) == dict
result = list(result.values())
return 1.0 - sum(result) / len(result)
def test_grace():
args = get_unsupervised_nn_args()
args.model = "grace"
args.num_layers = 2
args.max_epoch = 2
args.drop_feature_rates = [0.1, 0.2]
args.drop_edge_rates = [0.2, 0.3]
args.activation = "relu"
args.proj_hidden_size = 32
args.tau = 0.5
args.dataset = "cora"
args, dataset = build_nn_dataset(args)
for bs in [-1, 512]:
args.batch_size = bs
task = build_task(args)
ret = task.train()
assert ret["Acc"] > 0
def test_pairnorm_cora_deepgat():
args = get_default_args()
args.task = "node_classification"
args.dataset = "cora"
args.model = "pairnorm"
args.pn_model = "DeepGAT"
args.nlayer = 10
args.missing_rate = 100
args.norm_mode = "PN-SI"
args.residual = 0
args.hidden_layers = 64
args.nhead = 2
args.dropout = 0.6
args.norm_scale = 1.0
args.no_fea_norm = "store_false"
task = build_task(args)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def test_hin2vec_dblp():
args = get_default_args()
args.task = 'multiplex_node_classification'
args.dataset = 'gtn-dblp'
args.model = 'hin2vec'
dataset = build_dataset(args)
args.walk_length = 5
args.walk_num = 1
args.negative = 3
args.batch_size = 1000
args.hop = 2
args.epoches = 1
args.lr = 0.025
args.cpu = True
model = build_model(args)
task = build_task(args)
ret = task.train()
assert ret['f1'] > 0
def test_gatne_amazon():
args = get_default_args()
args.task = "multiplex_link_prediction"
args.dataset = "amazon"
args.model = "gatne"
args.walk_length = 5
args.walk_num = 1
args.window_size = 3
args.worker = 5
args.schema = None
args.epoch = 1
args.batch_size = 128
args.edge_dim = 5
args.att_dim = 5
args.negative_samples = 5
args.neighbor_samples = 5
task = build_task(args)
ret = task.train()
assert ret["ROC_AUC"] >= 0 and ret["ROC_AUC"] <= 1
def test_sdne_ppi():
args = get_default_args()
args.task = 'link_prediction'
args.dataset = 'ppi'
args.model = 'sdne'
dataset = build_dataset(args)
args.hidden_size1 = 100
args.hidden_size2 = 16
args.droput = 0.2
args.alpha = 0.01
args.beta = 5
args.nu1 = 1e-4
args.nu2 = 1e-3
args.max_epoch = 1
args.lr = 0.001
model = build_model(args)
task = build_task(args)
ret = task.train()
assert ret['ROC_AUC'] >= 0 and ret['ROC_AUC'] <= 1
def test_graph_mix():
args = get_default_args()
args.task = "node_classification"
args.dataset = "cora"
args.model = "gcnmix"
dataset = build_dataset(args)
args.num_features = dataset.num_features
args.num_classes = dataset.num_classes
args.rampup_starts = 1
args.rampup_ends = 100
args.mixup_consistency = 5.0
args.ema_decay = 0.999
args.alpha = 1.0
args.temperature = 1.0
args.k = 10
model = build_model(args)
task = build_task(args)
ret = task.train()
assert ret["Acc"] >= 0 and ret["Acc"] <= 1
def test_sdne_ppi():
args = get_default_args()
args.task = 'unsupervised_node_classification'
args.dataset = 'ppi'
args.model = 'sdne'
dataset = build_dataset(args)
args.hidden_size1 = 100
args.hidden_size2 = 16
args.droput = 0.2
args.alpha = 0.01
args.beta = 5
args.nu1 = 1e-4
args.nu2 = 1e-3
args.max_epoch = 1
args.lr = 0.001
model = build_model(args)
task = build_task(args)
ret = task.train()
assert ret['Micro-F1 0.9'] > 0
def test_lightgcn_ali():
args = get_default_args()
args.dataset = "ali"
args.model = "lightgcn"
args.Ks = [1]
args.dim = 8
args.l2 = 1e-4
args.mess_dropout = False
args.mess_dropout_rate = 0.0
args.edge_dropout = False
args.edge_dropout_rate = 0.0
args.ns = "rns"
args.K = 1
args.n_negs = 1
args.pool = "mean"
args.context_hops = 1
task = build_task(args)
ret = task.train(unittest=True)
assert ret["Recall"] >= 0
def test_jknet_jknet_cora():
args = get_default_args()
args.task = "node_classification"
args.dataset = "cora"
args.model = "jknet"
args.lr = 0.005
args.layer_aggregation = "maxpool"
args.node_aggregation = "sum"
args.n_layers = 3
args.n_units = 16
args.in_features = 1433
args.out_features = 7
args.max_epoch = 2
for aggr in ["maxpool", "concat"]:
args.layer_aggregation = aggr
task = build_task(args)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def test_pyg_gcn_cora_sampler():
args = get_default_args()
args.task = "node_classification"
args.dataset = "cora"
args.model = "pyg_gcn"
args.cpu = True
args.num_layers = 2
args.sample_coverage = 20
args.size_subgraph = 200
args.num_walks = 20
args.walk_length = 10
args.size_frontier = 20
sampler_list = ["node", "edge", "rw", "mrw"]
for sampler in sampler_list:
args.sampler = sampler
task = build_task(args)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def test_gdc_gcn_cora():
args = get_default_args()
args.task = "node_classification"
args.dataset = "cora"
args.model = "gdc_gcn"
dataset = build_dataset(args)
args.num_features = dataset.num_features
args.num_classes = dataset.num_classes
args.num_layers = 1
args.alpha = 0.05 # ppr filter param
args.t = 5.0 # heat filter param
args.k = 128 # top k entries to be retained
args.eps = 0.01 # change depending on gdc_type
args.dataset = dataset
args.gdc_type = "ppr" # ppr, heat, none
model = build_model(args)
task = build_task(args, dataset=dataset, model=model)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def test_pprgo_cora():
args = get_default_args()
args.cpu = True
args.task = "node_classification"
args.dataset = "cora"
args.model = "pprgo"
args.k = 32
args.alpha = 0.5
args.eval_step = 1
args.batch_size = 32
args.test_batch_size = 128
args.activation = "relu"
args.num_layers = 2
args.nprop_inference = 2
args.eps = 0.001
for norm in ["sym", "row"]:
args.norm = norm
task = build_task(args)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def test_dropedge_inceptiongcn_cora():
args = get_default_args()
args.task = "node_classification"
args.dataset = "cora"
args.model = "dropedge_gcn"
args.baseblock = "inceptiongcn"
args.inputlayer = "gcn"
args.outputlayer = "gcn"
args.hidden_size = 64
args.dropout = 0.5
args.withbn = False
args.withloop = False
args.nhiddenlayer = 1
args.nbaseblocklayer = 1
args.aggrmethod = "add"
args.activation = F.relu
args.task_type = "full"
task = build_task(args)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def test_gatne_amazon():
args = get_default_args()
args.task = 'multiplex_link_prediction'
args.dataset = 'amazon'
args.model = 'gatne'
dataset = build_dataset(args)
args.walk_length = 5
args.walk_num = 1
args.window_size = 3
args.worker = 5
args.iteration = 1
args.epoch = 1
args.batch_size = 128
args.edge_dim = 5
args.att_dim = 5
args.negative_samples = 5
args.neighbor_samples = 5
model = build_model(args)
task = build_task(args)
ret = task.train()
assert ret['ROC_AUC'] >= 0 and ret['ROC_AUC'] <= 1
def test_graphsaint_cora():
args = get_default_args()
args.task = "node_classification"
args.dataset = "cora"
args.trainer = "graphsaint"
args.model = "graphsaint"
args.valid_cpu = True
args.cpu = True
args.architecture = "1-1-0"
args.aggr = "concat"
args.act = "relu"
args.bias = "norm"
args.sample_coverage = 10
args.size_subgraph = 200
args.num_walks = 20
args.walk_length = 10
args.size_frontier = 20
args.sampler = "node"
task = build_task(args)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def train():
args = build_default_args()
dataset, args = get_dataset(args)
combinations = get_parameters()
best_parameters = None
best_result = None
best_val_acc = 0
print(f"===== Start At: {get_time()} ===========")
start = time.time()
random_seeds = list(range(5))
for item in combinations:
for key, val in item.items():
setattr(args, key, val)
print(f"### -- Parameters: {args.__dict__}")
result_list = []
for seed in random_seeds:
set_random_seed(seed)
task = build_task(args, dataset=dataset)
res = task.train()
result_list.append(res)
val_acc = [x["ValAcc"] for x in result_list]
test_acc = [x["Acc"] for x in result_list]
val_acc = sum(val_acc) / len(val_acc)
print(f"### Result: {val_acc}")
if val_acc > best_val_acc:
best_parameters = copy.deepcopy(args)
best_result = dict(Acc=sum(test_acc) / len(test_acc),
ValAcc=val_acc)
print(f"Best Parameters: {best_parameters}")
print(f"Best result: {best_result}")
end = time.time()
print(f"===== End At: {get_time()} ===========")
print("Time cost:", end - start)
def test_supergat():
args = get_default_args()
args.model = "supergat"
args.trainer = None
args.heads = 8
args.attention_type = "mask_only"
args.neg_sample_ratio = 0.5
args.edge_sampling_ratio = 0.8
args.val_interval = 1
args.att_lambda = 10
args.pretraining_noise_ratio = 0
args.to_undirected_at_neg = False
args.to_undirected = False
args.out_heads = None
args.total_pretraining_epoch = 0
args.super_gat_criterion = None
args.scaling_factor = None
dataset = build_dataset(args)
model = build_model(args)
task = build_task(args, dataset=dataset, model=model)
ret = task.train()
assert 0 <= ret["Acc"] <= 1
def test_deepergcn_cora():
args = get_default_args()
args.dataset = "cora"
args.task = "node_classification"
args.model = "deepergcn"
args.num_layers = 2
args.connection = "res+"
args.cluster_number = 3
args.learn_beta = True
args.learn_msg_scale = True
args.aggr = "softmax_sg"
args.batch_size = 1
args.activation = "relu"
args.beta = 1.0
args.p = 1.0
args.use_msg_norm = True
args.learn_p = True
args.learn_beta = True
args.learn_msg_scale = True
task = build_task(args)
ret = task.train()
assert 0 < ret["Acc"] < 1
def __init__(self, app: str, model: str, **kwargs):
super(RecommendationPipepline, self).__init__(app,
model=model,
**kwargs)
if "data" in kwargs:
data = kwargs["data"]
val_data = test_data = data[-100:, :]
data = build_recommendation_data("custom", data, val_data,
test_data)
self.data_path = kwargs.get("data_path", "tmp_data.pt")
self.batch_size = kwargs.get("batch_size", 128)
torch.save(data, self.data_path)
self.dataset = NodeDataset(path=self.data_path, scale_feat=False)
elif "dataset" in kwargs:
dataset = kwargs.pop("dataset")
self.dataset = build_dataset_from_name(dataset)
else:
print("Please provide recommendation data!")
exit(0)
self.batch_size = kwargs.get("batch_size", 2048)
self.n_items = self.dataset[0].n_params["n_items"]
args = get_default_args(task="recommendation",
dataset="ali",
model=model,
**kwargs)
args.model = args.model[0]
task = build_task(args, dataset=self.dataset)
task.train()
self.model = task.model
self.model.eval()
self.user_emb, self.item_emb = self.model.generate()
def test_stpgnn_infomax():
args = get_strategies_for_pretrain_args()
args.pretrain_task = "infomax"
task = build_task(args)
ret = task.train()
assert 0 < ret["Acc"] <= 1
def test_gcn_cora():
args = get_gnn_link_prediction_args()
print(args.evaluate_interval)
task = build_task(args)
ret = task.train()
assert 0.5 <= ret["AUC"] <= 1.0
from cogdl.utils import build_args_from_dict
def get_default_args():
cuda_available = torch.cuda.is_available()
default_dict = {
'hidden_size': 16,
'dropout': 0.5,
'patience': 100,
'max_epoch': 500,
'cpu': not cuda_available,
'lr': 0.01,
'device_id': [0],
'weight_decay': 5e-4
}
return build_args_from_dict(default_dict)
if __name__ == "__main__":
args = get_default_args()
args.task = 'node_classification'
args.dataset = 'cora'
args.model = 'gcn'
dataset = build_dataset(args)
args.num_features = dataset.num_features
args.num_classes = dataset.num_classes
args.num_layers = 2
model = build_model(args)
task = build_task(args, dataset=dataset, model=model)
ret = task.train()
from cogdl.tasks import build_task
from cogdl.options import get_default_args
args = get_default_args(task="node_classification",
dataset="cora",
model="gcn")
task = build_task(args)
ret = task.train()