timesead.data.statistics
========================

.. py:module:: timesead.data.statistics


Functions
---------

.. autoapisummary::

   timesead.data.statistics.get_data_all
   timesead.data.statistics.compute_whiten_matrix
   timesead.data.statistics.compute_feature_statistics
   timesead.data.statistics.compute_feature_mean_std
   timesead.data.statistics.compute_ts_statisitcs
   timesead.data.statistics.training_means
   timesead.data.statistics.compute_anomaly_positions
   timesead.data.statistics.compute_anomaly_lengths
   timesead.data.statistics.compute_total_time_steps


Module Contents
---------------

.. py:function:: get_data_all(dataset: torch.utils.data.Dataset, take_fd: bool = False) -> torch.Tensor

   Extract data matrix from dataset.

   :param dataset: Time-series dataset.
          take_fd: flag, to take first differences.
   :type dataset: torch.utils.data.Dataset
         take_id: bool
   :return: data matrix
   :rtype: torch.Tensor


.. py:function:: compute_whiten_matrix(dataset: torch.utils.data.Dataset, take_fd: bool = False) -> torch.Tensor

   Compute whiten matrix of dataset.

   :param dataset: Time-series dataset.
   :type dataset: torch.utils.data.Dataset
   :return: whiten matrix
   :rtype: torch.Tensor


.. py:function:: compute_feature_statistics(dataset: torch.utils.data.Dataset, take_fd: bool = False) -> Tuple[Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], torch.Tensor, torch.Tensor]

   Compute statistics of each feature.

   :param dataset: Time-series dataset.
   :type dataset: torch.utils.data.Dataset
   :return: Normal mean, normal std, anomaly mean, anomaly std, minimum, and maximum
   :rtype: Tuple[Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], torch.Tensor, torch.Tensor]


.. py:function:: compute_feature_mean_std(dataset: timesead.data.dataset.BaseTSDataset, take_fd: bool = False) -> Tuple[Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor]]

   Compute the mean and standard deviation of each feature.
   :param dataset: Time-series dataset.
   :type dataset: torch.utils.data.Dataset
   :return: Normal mean, normal std, anomaly mean, anomaly std
   :rtype: Tuple[Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor]]


.. py:function:: compute_ts_statisitcs(timeseries: torch.Tensor, targets: torch.Tensor, dim_order: str = 'tf') -> Tuple[Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], Optional[torch.Tensor], torch.Tensor, torch.Tensor]

.. py:function:: training_means(dataset)

.. py:function:: compute_anomaly_positions(dataset: torch.utils.data.Dataset) -> List[int]

   Computes the positions of anomalies in the dataset.

   :param dataset: Dataset to compute the statistics of.
   :type dataset: torch.utils.data.Dataset
   :return: List of all the relative positions of anomalies in the dataset.
   :rtype: List[int]


.. py:function:: compute_anomaly_lengths(dataset: torch.utils.data.Dataset) -> List[int]

   Computes the length of each anomalous window in the dataset.

   :param dataset: Dataset to compute the statistics of.
   :type dataset: torch.utils.data.Dataset
   :return: List of lengths of anomalies in the dataset.
   :rtype: List[int]


.. py:function:: compute_total_time_steps(dataset: torch.utils.data.Dataset) -> int

   Compute the total amount of time steps in the dataset (normal + anormal)

   :param dataset: Dataset to compute the statistics of.
   :type dataset: torch.utils.data.Dataset
   :return: Number of time steps in the dataset
   :rtype: int