def loadTrainData(subjects): trainData = {} completeData = [] for subj in subjects: for action in actions: for subact in subactions: #filename = '../../../dataset' filename = '{0}/{1}/{2}_{3}.txt'.format( path_to_dataset, subj, action, subact) action_sequence = readCSVasFloat(filename) T = action_sequence.shape[0] odd_list = range(1, T, 2) even_list = range(0, T, 2) trainData[(subj, action, subact)] = action_sequence trainData[(subj, action, subact, 'even')] = action_sequence[even_list, :] trainData[(subj, action, subact, 'odd')] = action_sequence[odd_list, :] if len(completeData) == 0: completeData = copy.deepcopy(trainData[(subj, action, subact)]) else: completeData = np.append(completeData, trainData[(subj, action, subact)], axis=0) return trainData, completeData
def convertAndSave(fname): fpath = path_to_trajfiles + fname if not os.path.exists(fpath): return False normalizedData = readCSVasFloat(fpath) origData = unNormalizeData(normalizedData,data_mean,data_std,dimensions_to_ignore) if len(origData) > 0: fpath = path_to_trajfiles + fname writeMatToCSV(origData,fpath) return True
def convertAndSave(fname): fpath = path_to_trajfiles + fname if not os.path.exists(fpath): return False normalizedData = readCSVasFloat(fpath) origData = unNormalizeData(normalizedData, data_mean, data_std, dimensions_to_ignore) if len(origData) > 0: fpath = path_to_trajfiles + fname writeMatToCSV(origData, fpath) return True
def loadTrainData(subjects): trainData = {} completeData = [] for subj in subjects: for action in actions: for subact in subactions: filename = "{0}/{1}/{2}_{3}.txt".format(path_to_dataset, subj, action, subact) action_sequence = readCSVasFloat(filename) T = action_sequence.shape[0] odd_list = range(1, T, 2) even_list = range(0, T, 2) trainData[(subj, action, subact)] = action_sequence trainData[(subj, action, subact, "even")] = action_sequence[even_list, :] trainData[(subj, action, subact, "odd")] = action_sequence[odd_list, :] if len(completeData) == 0: completeData = copy.deepcopy(trainData[(subj, action, subact)]) else: completeData = np.append(completeData, trainData[(subj, action, subact)], axis=0) return trainData, completeData