RobustScaler

Scale features using statistics that are robust to outliers.

This Scaler removes the median and scales the data according to the interquantile range.

Parameters

• with_centering

  Default → True

  Whether to centre the data before scaling.

• with_scaling

  Default → True

  Whether to scale data to IQR.

• q_inf

  Default → 0.25

  Desired inferior quantile, must be between 0 and 1.

• q_sup

  Default → 0.75

  Desired superior quantile, must be between 0 and 1.

Attributes

• median (dict)

  Mapping between features and instances of stats.Quantile(0.5)`.

• iqr (dict)

  Mapping between features and instances of stats.IQR.

Examples

from pprint import pprint
import random
from river import preprocessing

random.seed(42)
X = [{'x': random.uniform(8, 12)} for _ in range(5)]
pprint(X)

[{'x': 10.557707},
    {'x': 8.100043},
    {'x': 9.100117},
    {'x': 8.892842},
    {'x': 10.945884}]

scaler = preprocessing.RobustScaler()

for x in X:
    scaler.learn_one(x)
    print(scaler.transform_one(x))

    {'x': 0.0}
    {'x': -1.0}
    {'x': 0.0}
    {'x': -0.12449923287875722}
    {'x': 1.1086595155704708}

Methods

learn_one

Update with a set of features x.

A lot of transformers don't actually have to do anything during the learn_one step because they are stateless. For this reason the default behavior of this function is to do nothing. Transformers that however do something during the learn_one can override this method.

Parameters

• x — 'dict'

transform_one

Transform a set of features x.

Parameters

• x — 'dict'

Returns

dict: The transformed values.