Esempi in Python per print_result

Esempio n. 1

@register_func("blogcatalog")
def blog_config(args):
    return args


@register_func("wikipedia")
def wiki_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["ppi", "blogcatalog", "wikipedia"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "netmf")

Esempio n. 2

File: pte.py Progetto: zhjhr181/cogdl

@register_func("gtn-acm")
def acm_config(args):
    return args


@register_func("gtn-imdb")
def imdb_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_multiplex_node_classification(dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["gtn-dblp", "gtn-acm", "gtn-imdb"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "pte")

Esempio n. 3

File: metapath2vec.py Progetto: zhjhr181/cogdl

@register_func("gtn-acm")
def acm_config(args):
    return args


@register_func("gtn-imdb")
def imdb_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_multiplex_node_classification(dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["gtn-dblp", "gtn-acm", "gtn-imdb"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "metapath2vec")

Esempio n. 4

@register_func("blogcatalog")
def blog_config(args):
    return args


@register_func("wikipedia")
def wiki_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["ppi", "blogcatalog", "wikipedia"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "sdne")

Esempio n. 5

File: base_example.py Progetto: zrt/cogdl

def register_func(name):
    def register_func_name(func):
        DATASET_REGISTRY[name] = func
        return func

    return register_func_name


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = []
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "model_name")

Esempio n. 6

@register_func("youtube")
def acm_config(args):
    return args


@register_func("twitter")
def imdb_config(args):
    args.eval_type = "1"
    return args


def run(dataset_name):
    args = build_default_args_for_multiplex_link_prediction(dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["amazon", "youtube", "twitter"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "gatne")

Esempio n. 7

File: hope.py Progetto: zrt/cogdl

@register_func("blogcatalog")
def blog_config(args):
    return args


@register_func("wikipedia")
def wiki_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["ppi", "blogcatalog", "wikipedia"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "hope")

Esempio n. 8

File: deepwalk.py Progetto: zhjhr181/cogdl

@register_func("blogcatalog")
def blog_config(args):
    return args


@register_func("wikipedia")
def wiki_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["ppi", "blogcatalog", "wikipedia"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "deepwalk")

Esempio n. 9

@register_func("blogcatalog")
def blog_config(args):
    return args


@register_func("wikipedia")
def wiki_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["ppi", "blogcatalog", "wikipedia"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "grarep")

Esempio n. 10

@register_func("gtn-imdb")
def imdb_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_heterogeneous_node_classification(dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    args.num_features = dataset.num_features
    args.num_classes = dataset.num_classes
    args.num_edge = dataset.num_edge
    args.num_nodes = dataset.num_nodes
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["gtn-dblp", "gtn-acm", "gtn-imdb"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "han")

Esempio n. 11

File: spectral.py Progetto: yaoxingcheng/cogdl

@register_func("blogcatalog")
def blog_config(args):
    return args


@register_func("wikipedia")
def wiki_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["ppi", "blogcatalog", "wikipedia"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "spectral")

Esempio n. 12

@register_func("blogcatalog")
def blog_config(args):
    return args


@register_func("wikipedia")
def wiki_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_unsupervised_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["ppi", "blogcatalog", "wikipedia"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "prone")

Esempio n. 13

File: hin2vec.py Progetto: zhjhr181/cogdl

@register_func("gtn-acm")
def acm_config(args):
    return args


@register_func("gtn-imdb")
def imdb_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_multiplex_node_classification(dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["gtn-dblp", "gtn-acm", "gtn-imdb"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "hin2vec")

Esempio n. 14

@register_func("gtn-imdb")
def imdb_config(args):
    return args


def run(dataset_name):
    args = build_default_args_for_heterogeneous_node_classification(
        dataset_name)
    args = DATASET_REGISTRY[dataset_name](args)
    dataset = build_dataset(args)
    args.num_features = dataset.num_features
    args.num_classes = dataset.num_classes
    args.num_edge = dataset.num_edge
    args.num_nodes = dataset.num_nodes
    results = []
    for seed in args.seed:
        set_random_seed(seed)
        task = build_task(args, dataset=dataset)
        result = task.train()
        results.append(result)
    return results


if __name__ == "__main__":
    datasets = ["gtn-dblp", "gtn-acm", "gtn-imdb"]
    results = []
    for x in datasets:
        results += run(x)
    print_result(results, datasets, "gtn")