Skip to content

Latest commit

 

History

History

aggregate

Aggregate


Performs a Rolling Aggregate on the input data according to specifications.

Given:

  • df_artifact: Input DataFrame
  • keys: Labels to work by
  • metrics: List of metrics to perform the aggregate task on
  • labels: List of labels to perform the label aggregate on
  • metric_aggs: Array of aggregations to run on the metrics
    ex: ['mean', 'sum']
  • label_aggs: Array of aggregations to run on the labels
    ex: ['max']
  • suffix: A suffix to add to the aggregations (ex: '_daily')
  • window: Window size, can be int for number of rows, or in time interval string if timestamp index is available
  • center: Perform aggregation on the sample at the window's center
  • inplace:
  • True: Returns only the newly aggregated data
  • False: Returns the newly aggregated data with the original dataset