PB preprocessed data used in paper "Multi variables time series information bottleneck"

Preprocessed PB data used in paper "Multi variables time series information bottleneck" with the GitHub code This dataset is created from a public available dataset of solar power data collected in Alabama by CalTrans. The npz file is a numpy (np) compressed data and can be loaded using np.load with allow_pickle=True<br> Loaded data is then a python dict described bellow. Each sample 'data' is a np.ndarray with 2 dimensions: time (various length) and wavelength (length=325 representing 325 traffic detectors ordered like in CalTrans). Each sample is given a 'position' which is a list of length 4:<br> position[1] is a string that gives the name of the event<br> position[4] is a boolean vector that gives the time positionsof the corresponding sample in the original sequence of public IRIS level2 data Data file info :<br> Type: .npz<br> Size: 114.23MB<br> *** Key: 'data_TR_PB'<br> ndarray data of length 3<br> containing np.ndarray of shapes [12160, 325] *** Key: 'data_VAL_PB'<br> ndarray data of length 3<br> containing np.ndarray of shapes [868, 325] *** Key: 'data_TE_PB'<br> ndarray data of length 3<br> containing np.ndarray of shapes [4343, 325] *** Key: 'data_TR'<br> ndarray data of length 3<br> containing np.ndarray of shapes [12160, 325] *** Key: 'data_VAL'<br> ndarray data of length 3<br> containing np.ndarray of shapes [868, 325] *** Key: 'data_TE'<br> ndarray data of length 3<br> containing np.ndarray of shapes [4343, 325] *** Key: 'position_TR_PB'<br> ndarray data of length 3<br> containing ndarray data of length 4<br> containing mix of types {'str', 'ndarray', 'int'} *** Key: 'position_VAL_PB'<br> ndarray data of length 3<br> containing ndarray data of length 4<br> containing mix of types {'str', 'ndarray', 'int'} *** Key: 'position_TE_PB'<br> ndarray data of length 3<br> containing ndarray data of length 4<br> containing mix of types {'str', 'ndarray', 'int'} *** Key: 'position_TR'<br> ndarray data of length 3<br> containing ndarray data of length 4<br> containing mix of types {'str', 'ndarray', 'int'} *** Key: 'position_VAL'<br> ndarray data of length 3<br> containing ndarray data of length 4<br> containing mix of types {'str', 'ndarray', 'int'} *** Key: 'position_TE'<br> ndarray data of length 3<br> containing ndarray data of length 4<br> containing mix of types {'str', 'ndarray', 'int'}本预处理PB数据用于论文《多变量时间序列信息瓶颈》（Multi variables time series information bottleneck），配套GitHub代码已公开。该数据集源自CalTrans（加利福尼亚州运输部）在阿拉巴马州采集的公开太阳能功率数据集。该npz文件为NumPy（np）压缩格式数据，可通过`np.load(allow_pickle=True)`加载。加载后的数据为Python字典，结构说明如下：每个样本的`data`字段为二维NumPy数组，维度分别为时间（长度不定）与波长（长度为325，对应CalTrans按顺序排列的325个交通检测器）。每个样本附带`position`字段，为长度为4的列表： - `position[1]`为字符串，用于表示事件名称； - `position[4]`为布尔向量，用于标记该样本在公开IRIS Level2数据原始序列中的时间位置。 数据文件详情如下： - 格式：.npz - 大小：114.23MB - 关键字段说明： 1. 'data_TR_PB'：长度为3的NumPy数组，内部包含形状为[12160, 325]的NumPy数组 2. 'data_VAL_PB'：长度为3的NumPy数组，内部包含形状为[868, 325]的NumPy数组 3. 'data_TE_PB'：长度为3的NumPy数组，内部包含形状为[4343, 325]的NumPy数组 4. 'data_TR'：长度为3的NumPy数组，内部包含形状为[12160, 325]的NumPy数组 5. 'data_VAL'：长度为3的NumPy数组，内部包含形状为[868, 325]的NumPy数组 6. 'data_TE'：长度为3的NumPy数组，内部包含形状为[4343, 325]的NumPy数组 7. 'position_TR_PB'：长度为3的NumPy数组，内部包含长度为4的NumPy数组，元素类型包含{'str', 'ndarray', 'int'} 8. 'position_VAL_PB'：长度为3的NumPy数组，内部包含长度为4的NumPy数组，元素类型包含{'str', 'ndarray', 'int'} 9. 'position_TE_PB'：长度为3的NumPy数组，内部包含长度为4的NumPy数组，元素类型包含{'str', 'ndarray', 'int'} 10. 'position_TR'：长度为3的NumPy数组，内部包含长度为4的NumPy数组，元素类型包含{'str', 'ndarray', 'int'} 11. 'position_VAL'：长度为3的NumPy数组，内部包含长度为4的NumPy数组，元素类型包含{'str', 'ndarray', 'int'} 12. 'position_TE'：长度为3的NumPy数组，内部包含长度为4的NumPy数组，元素类型包含{'str', 'ndarray', 'int'}