Ts ( pandas Series or DataFrame) – Time series to be transformed. Train models in pipeline sequentially, and transform time series Pandas Series, pandas DataFrame, list, or dict fit_transform ( ts, skip_fit=None, return_intermediate=False ) ¶ Pipeline, and use the last detector to detect anomalies. Train models in pipeline sequentially, transform time series along Pandas Series, pandas DataFrame, or dict fit_detect ( ts, skip_fit=None, return_intermediate=False, return_list=False ) ¶ If return_intermediate=False, return transformed series, i.e. Ts ( pandas Series or DataFrame) – Time series to be transformed Transform time series sequentially along pipeline. Pandas Series, pandas DataFrame, list, or dict transform ( ts, return_intermediate=False ) ¶ It is instantaneous or a 2-tuple of pandas Timestamps if it is a Will be a list of events where an event is a pandas Timestamp if If return_list=True, result from a detector or an aggregator Will be a binary pandas Series indicating normal/anomalous. If return_list=False, result from a detector or an aggregator
Pipeline as a dict where each item represents the result of a If return_intermediate=True, return results of all models in If return_intermediate=False, return detected anomalies, i.e. Return_list ( bool, optional) – Whether to return a list of anomalous events, or a binary series Ts ( pandas Series or DataFrame) – Time series to detect anomalies from. Transform time series sequentially along pipeline, and detectĪnomalies with the last detector. Step does not perform transformation or detection, the result ofĭict, optional detect ( ts, return_intermediate=False, return_list=False ) ¶ If return_intermediate=True, return intermediate results generatedĭuring training as a dictionary where keys are step names. Return_intermediate ( bool, optional) – Whether to return intermediate results. Models that are already trained by the same time series, and re. This could be used when pipeline contains Skip_fit ( list, optional) – Models to skip training. Ts ( pandas Series or DataFrame) – Time series used to train models. Train all models in the pipeline sequentially. steps = > myPipeline = Pipeline ( steps ) fit ( ts, skip_fit=None, return_intermediate=False ) ¶