def main(): try: args = get_args() data = preprocess(args) if args.visualize_data: visualize(data) sys.exit(1) train_set, test_set = split(data) num_examples = train_set.shape[0] num_features = train_set.shape[1] - 1 if args.mini_batch: batch_size = 32 # or 64 epochs = 1500 else: batch_size = num_examples epochs = 30000 nn = NeuralNetwork(num_features, batch_size, epochs) if args.train: nn.train(data, train_set, test_set, num_examples, args.quiet) if args.evaluation: y_pred = probability_to_class( nn.output.T) get_validation_metrics(y_pred[:, 0], nn.y.T[:, 0]) # mini-batch learning is noisy, so we don't plot it if not args.mini_batch: plot_learning(nn.train_losses, nn.test_losses) # save network params if args.save_model: W1, W2, W3, W4 = nn.weights1.tolist(), nn.weights2.tolist(), nn.weights3.tolist(), nn.weights4.tolist() B1, B2, B3, B4 = nn.bias1.tolist(), nn.bias2.tolist(), nn.bias3.tolist(), nn.bias4.tolist() model = dict(weights1=W1, weights2=W2, weights3=W3, weights4=W4, bias1=B1, bias2=B2, bias3=B3, bias4=B4) with open("model.json", "w") as f: json.dump(model, f, separators=(',', ':'), indent=4) if args.predict and (args.predict == "model.json"): try: with open(args.predict) as file: model = json.load(file) except: error_exit("please provide a valid model") nn.load_model(model) nn.predict(test_set, epochs) except: pass
def main(): # 获取命令行参数 command_args = get_args() # 设置logging日志级别 if command_args['debug'] == True: logging.getLogger().setLevel(logging.DEBUG) else: logging.getLogger().setLevel(logging.INFO) # 获取配置文件参数 file_opts = get_file_opts(command_args) # 实例化配置 collector_opts = CollectorConfig(file_opts=file_opts, command_args=command_args) # 实例化AliyunRDSCollector aliyun_rds_collector = AliyunRDSCollector(config=collector_opts) # 注册到Prometheus的registry里面 REGISTRY.register(aliyun_rds_collector) app = make_wsgi_app() httpd = make_server( host=str(collector_opts.server['host']), port=int(collector_opts.server['port']), app=app, ) logging.info("Start exporter, listen on {}:{}".format(str(collector_opts.server['host']), int(collector_opts.server['port']))) httpd.serve_forever()
confidences = t.split(';')[1].split( ' ') # ['', 'x_confs', '0.5362149'] confidences = confidences[2:] for c in confidences: confidences_sum += float(c) tot += 1 mean = confidences_sum / tot if tot != 0 else 0 confidences_per_line[cpt] = mean cpt += 1 return confidences_per_line if __name__ == '__main__': # Options parser options = tools.get_args() dir_path = options.data_dir level = options.level # For each file in the html directory for file in glob.glob(dir_path + '*.html'): # Parsing inFile = io.open(file, mode='r', encoding='utf-8') html = inFile.read() soup = BeautifulSoup(html, 'lxml') # parsing if level == 'page': doc_name = file.split('/')[-1] sum_, tot = get_confidences__page_level(soup) mean = sum_ / tot if tot != 0 else 0 print(doc_name + '\t' + str(mean)) elif level == 'line': doc_name = file.split('/')[-1]
if cpt_proc % 100 == 0: sys.stdout.write("\r%i processed, %i relevant" % (cpt_proc, cpt_rel)) # sys.stdout.flush() # if the buffer gets big, which is not the case output_path = write_output(output_dic, options) list_docs_not_found(missing_docs) return cpt_proc, cpt_rel, output_path if __name__ == "__main__": start = time.clock() options = get_args() print(options) if options.corpus == None: print( "Please specify a Json file (-c), see README.txt for more informations about the format. To use the default example :\n -c docs/Indonesian_GL.json" ) exit() else: options.document_path = "None" try: os.makedirs("tmp") except: pass cpt_doc, cpt_rel, output_path = start_detection(options) end = time.clock() print("\n%s docs proc. in %s seconds" %
return self.bhx_net.encode(state) def obs_encode(self, obs, hx=None): if hx is None: hx = Variable(torch.zeros(1, self.gru_units)) if next(self.parameters()).is_cuda: hx = hx.cuda() _, _, _, (_, _, _, input_x) = self.gru_net((obs, hx), input_fn=self.obx_net, hx_fn=self.bhx_net, inspect=True) return input_x if __name__ == '__main__': args = tl.get_args() env = gym.make(args.env) env.seed(args.env_seed) obs = env.reset() # create directories to store results result_dir = tl.ensure_directory_exits( os.path.join(args.result_dir, 'Classic_Control')) env_dir = tl.ensure_directory_exits(os.path.join(result_dir, args.env)) gru_dir = tl.ensure_directory_exits( os.path.join(env_dir, 'gru_{}'.format(args.gru_size))) gru_net_path = os.path.join(gru_dir, 'model.p') gru_plot_dir = tl.ensure_directory_exits(os.path.join(gru_dir, 'Plots')) bhx_dir = tl.ensure_directory_exits(
logging.basicConfig( level=logging.INFO, format=fmt, datefmt=datefmt, ) if filename_log: build_log = logging.FileHandler(filename=filename_log, mode='w') build_log.setLevel(logging.INFO) formatter = logging.Formatter(fmt, datefmt=datefmt) build_log.setFormatter(formatter) logging.getLogger('').addHandler(build_log) if __name__ == '__main__': args = get_args() logger.info(f"Reading test file") test_file = os.path.join(args.input_path, args.test_filename) test = process_data_classification(test_file) X_test = test.drop('AlexLabel', axis=1) texts_test = X_test["Text-EN"].str.lower() cards_test = X_test.filter( regex="card_.*" ) # Using a regular expression cause some cards may not be in test. y_test = test['AlexLabel'] logger.info(f" Using test dataset with {X_test.shape[0]} instances") logger.info(f"Reading train file") train_file = os.path.join(args.input_path, args.train_filename) train = process_data_classification(train_file) X_train = train.drop('AlexLabel', axis=1)
name = 'RandomForest' if args.no_ppi: name += '_NO-PPI' if args.expression: name += '_EXP' if args.sublocs: name += '_SUB' if args.orthologs: name += '_ORT' return name if __name__ == '__main__': args = tools.get_args() mean, std = main(args) name = get_name(args) df_path = 'results/results.csv' df = pd.read_csv(df_path) df.loc[len(df)] = [ name, args.organism, args.ppi, args.expression, args.orthologs, args.sublocs, args.n_runs, mean, std ] df.to_csv(df_path, index=False) print(df.head())