config.py

import configparser


class Config(object):
    def __init__(self, config_file):
        conf = configparser.ConfigParser()
        try:
            conf.read(config_file)
        except:
            print('loading config: %s failed' %(config_file))
        self.train_graph_file = conf.get("Graph_Data", "train_graph_file")
        if conf.has_option("Graph_Data", "origin_graph_file"):
            self.origin_graph_file = conf.get("Graph_Data", "origin_graph_file")
        else:
            self.origin_graph_file = False
        if conf.has_option("Graph_Data", "label_file"):
            self.label_file = conf.get("Graph_Data", "label_file")
        else:
            self.label_file = False

        if conf.has_option("Model_Setup", "restore_model"):
            self.restore_model = conf.get("Model_Steup", "restore_model")
        else:
            self.restore_model = False

        self.display = conf.getint("Output", "display")
        self.embedding_filename = conf.get("Output", "embedding_filename")
        if conf.has_option("Output", "check_reconstruction"):
            self.check_reconstruction = [int(i) for i in conf.get("Output", "check_reconstruction").split(',')]
        else:
            self.check_reconstruction = False
        if conf.has_option("Output", "check_link_prediction"):
            self.check_link_prediction = [int(i) for i in conf.get("Output", "check_link_prediction").split(',')]
        else:
            self.check_link_prediction = False
        if conf.has_option("Output", "check_classification"):
            self.check_classification = True
        else:
            self.check_classification = False

        self.struct = [int(i) for i in conf.get("Model_Setup", "struct").split(',')]
        self.alpha = conf.getfloat("Model_Setup", "alpha")
        self.gamma = conf.getfloat("Model_Setup", "gamma")
        self.reg = conf.getfloat("Model_Setup", "reg")
        self.beta = conf.getfloat("Model_Setup", "beta")

        self.batch_size = conf.getint("Model_Setup", "batch_size")
        self.epochs_limit = conf.getint("Model_Setup", "epochs_limit")
        self.learning_rate = conf.getfloat("Model_Setup", "learning_rate")

        self.DBN_init = conf.getboolean("Model_Setup", "dbn_init")
        self.dbn_epochs = conf.getint("Model_Setup", "dbn_epochs")
        self.dbn_batch_size = conf.getint("Model_Setup", "dbn_batch_size")
        self.dbn_learning_rate = conf.getfloat("Model_Setup", "dbn_learning_rate")

        self.spare_dot = False
        self.ng_sample_ratio = conf.getfloat("Model_Setup", "ng_sample_ratio")

        self.walk_times = conf.getint("Walk_Setup", "walk_times")
        self.walk_length = conf.getint("Walk_Setup", "walk_length")
        self.restart_ratio = conf.getfloat("Walk_Setup", "restart_ratio")

        self.T = conf.getfloat("Walk_Setup", "T")
        self.tao = conf.getfloat("Walk_Setup", "tao")