Skip to content


A linear detrender which centers the target in zero.

At each learn_one step, the current mean of y is subtracted from y before being fed to the provided regression model. During the predict_one step, the current mean is added to the prediction of the regression model.


  • regressor (base.Regressor)

  • window_size (int) – defaults to None

    Window size used for calculating the rolling mean. If None, then a mean over the whole target data will instead be used.



Return a fresh estimator with the same parameters.

The clone has the same parameters but has not been updated with any data. This works by looking at the parameters from the class signature. Each parameter is either - recursively cloned if it's a River classes. - deep-copied via copy.deepcopy if not. If the calling object is stochastic (i.e. it accepts a seed parameter) and has not been seeded, then the clone will not be idempotent. Indeed, this method's purpose if simply to return a new instance with the same input parameters.


Fits to a set of features x and a real-valued target y.


  • x (dict)
  • y (numbers.Number)


Regressor: self


Predicts the target value of a set of features x.


  • x (dict)


Number: The prediction.