parser.add_argument('--file', dest="file", type=str, default="RESULTS/football-5D-KMEANS-1/",
help="embeddings location file")
args = parser.parse_args()
dataset_dict = { "karate": corpora.load_karate,
"football": corpora.load_football,
"flickr": corpora.load_flickr,
"dblp": corpora.load_dblp,
"books": corpora.load_books,
"blogCatalog": corpora.load_blogCatalog,
"polblog": corpora.load_polblogs,
"adjnoun": corpora.load_adjnoun
}
log_in = logger.JSONLogger(os.path.join(args.file,"log.json"), mod="continue")
dataset_name = log_in["dataset"]
print(dataset_name)
n_gaussian = log_in["n_gaussian"]
if(dataset_name not in dataset_dict):
print("Dataset " + dataset_name + " does not exist, please select one of the following : ")
print(list(dataset_dict.keys()))
quit()
print("Loading Corpus ")
D, X, Y = dataset_dict[dataset_name]()
results = []
std_kmeans = []
representations = torch.load(os.path.join(args.file,"embeddings.t7"))[0]
if (args.init_beta < 0):
args.init_beta = args.beta
# set the seed for random sampling
alpha, beta = args.init_alpha, args.init_beta
print("Loading Corpus ")
D, X, Y = dataset_dict[args.dataset]()
print("Creating dataset")
# index of examples dataset
dataset_index = corpora_tools.from_indexable(
torch.arange(0, len(D), 1).unsqueeze(-1))
print("Dataset Size -> ", len(D))
if (args.save):
os.makedirs(os.path.join(saving_folder, args.id + "/"), exist_ok=True)
logger_object = logger.JSONLogger(
os.path.join(saving_folder, args.id + "/log.json"))
logger_object.append(vars(args))
D.set_path(False)
# negative sampling distribution
frequency = D.getFrequency()**(3 / 4)
frequency[:, 1] /= frequency[:, 1].sum()
frequency = pytorch_categorical.Categorical(frequency[:, 1])
# random walk dataset
d_rw = D.light_copy()
rw_log = logger.JSONLogger("ressources/random_walk.conf", mod="continue")
if (args.force_rw):
key = args.dataset + "_" + str(args.context_size) + "_" + str(
args.walk_lenght) + "_" + str(args.seed)
if (key in rw_log):
print("Creating dataset")
# index of examples dataset
dataset_index = corpora_tools.from_indexable(
torch.arange(0, len(D), 1).unsqueeze(-1))
print("Dataset Size -> ", len(D))
D.set_path(False)
# negative sampling distribution
frequency = D.getFrequency()**(3 / 4)
frequency[:, 1] /= frequency[:, 1].sum()
frequency = pytorch_categorical.Categorical(frequency[:, 1])
# random walk dataset
d_rw = D.light_copy()
rw_log = logger.JSONLogger("ressources/random_walk.conf", mod="continue")
if (args.force_rw):
key = args.dataset + "_" + str(args.context_size) + "_" + str(
args.walk_lenght) + "_" + str(args.seed)
if (key in rw_log):
try:
print('Loading random walks from files')
d_rw = torch.load(rw_log[key]["file"])
print('Loaded')
except:
os.makedirs("/local/gerald/KMEANS_RESULTS/", exist_ok=True)
d_rw.set_walk(args.walk_lenght, 1.0)
d_rw.set_path(True)
d_rw = corpora.ContextCorpus(d_rw,
context_size=args.context_size,
"dblp": corpora.load_dblp,
"books": corpora.load_books,
"blogCatalog": corpora.load_blogCatalog
}
optimizer_dict = {
"addhsgd": optimizer.PoincareBallSGDAdd,
"exphsgd": optimizer.PoincareBallSGDExp,
"hsgd": optimizer.PoincareBallSGD,
"exphsga": optimizer.PoincareBallSGAExp
}
if (args.save):
print("The following options are use for the current experiment ", args)
os.makedirs("RESULTS/" + args.id + "/", exist_ok=True)
logger_object = logger.JSONLogger("RESULTS/" + args.id + "/log.json")
logger_object.append(vars(args))
# check if dataset exists
if (args.dataset not in dataset_dict):
print("Dataset " + args.dataset +
" does not exist, please select one of the following : ")
print(list(dataset_dict.keys()))
quit()
if (args.embedding_optimizer not in optimizer_dict):
print("Optimizer " + args.embedding_optimizer +
" does not exist, please select one of the following : ")
print(list(optimizer_dict.keys()))
quit()