def get_input_data(self, skip_preprocessing, preprocessing_constant, normalization_method, database_path, feature_names, datetime_interval, blockchain_indicators): data = md.get_dataset_with_descriptors(skip_preprocessing = skip_preprocessing, preproc_constant = preprocessing_constant, normalization_method = normalization_method, dataset_directory = database_path, feature_names = feature_names, datetime_interval = datetime_interval, blockchain_indicators = blockchain_indicators) self.data = data self.feature_names = feature_names return self.data
def plot_vertical_lines(data, desired_color='brown'): ## even data samples data1 = [data[i] if ((i % 2) == 0) else 0 for i in range(len(data))] #odd data samples data2 = [data[i] if ((i % 2) != 0) else 0 for i in range(len(data))] #plot them plt.plot(data1, color=desired_color) plt.plot(data2, color=desired_color) directory = '/home/catalin/databases/klines_2014-2018_15min/' hard_coded_file_number = 0 data = md.get_dataset_with_descriptors( concatenate_datasets_preproc_flag=True, preproc_constant=0.99, normalization_method="rescale", dataset_directory=directory, hard_coded_file_number=hard_coded_file_number) X = data['preprocessed_data'] ## this will be used for training X_unprocessed = data['data'] close_prices = X_unprocessed[:, 0] one_day_in_15_min_candles = 15 * 4 * 24 twelve_hrs_in_15_min_candles = 15 * 4 * 12 six_hrs_in_15_min_candles = 15 * 4 * 6 three_hrs_in_15_min_candles = 15 * 4 * 3 one_hr_in_15_min_candles = 15 * 4 SMA_12_day_values, EMA_12_day_values = SMA_EMA(close_prices, 12 * 1) SMA_26_day_values, EMA_26_day_values = SMA_EMA(close_prices, 26 * 1)
# 'slope_VBP_smooth_24', 'nlms_indicator', 'nlms_smoothed_indicator', # 'rls_indicator_error', # 'rls_smoothed_indicator', 'ATR_EMA', 'ATR_EMA_Wilder', 'CMF_12h', 'CMF_12h_2', # 'sentiment_indicator_positive', # 'sentiment_indicator_negative' ] data = md.get_dataset_with_descriptors(skip_preprocessing=False, preproc_constant=0.99, normalization_method="rescale", dataset_directory=directory, feature_names=feature_names) X = data['preprocessed_data'] ## this will be used for training X_unprocessed = data['data'] #sys.exit() start_date = md.get_date_from_UTC_ms(data['dataset_dict']['UTC'][0]) end_date = md.get_date_from_UTC_ms(data['dataset_dict']['UTC'][-1]) #import random # #noise = np.array([np.array([random.uniform(0,1) for _ in range(X.shape[0])]) for i in range(5)]) # #X = np.concatenate([X, noise.T], axis = 1) #X_unprocessed = np.concatenate([X_unprocessed, noise.T], axis = 1)