RobustScaler

class cuml.preprocessing.RobustScaler(*args, **kwargs)

Scale features using statistics that are robust to outliers.

This Scaler removes the median and scales the data according to the quantile range (defaults to IQR: Interquartile Range). The IQR is the range between the 1st quartile (25th quantile) and the 3rd quartile (75th quantile).

Centering and scaling happen independently on each feature by computing the relevant statistics on the samples in the training set. Median and interquartile range are then stored to be used on later data using the transform method.

Standardization of a dataset is a common requirement for many machine learning estimators. Typically this is done by removing the mean and scaling to unit variance. However, outliers can often influence the sample mean / variance in a negative way. In such cases, the median and the interquartile range often give better results.

Parameters:

with_centeringboolean, default=True

If True, center the data before scaling. This will cause transform to raise an exception when attempted on sparse matrices, because centering them entails building a dense matrix which in common use cases is likely to be too large to fit in memory.

with_scalingboolean, default=True

If True, scale the data to interquartile range.

quantile_rangetuple (q_min, q_max), 0.0 < q_min < q_max < 100.0

Default: (25.0, 75.0) = (1st quantile, 3rd quantile) = IQR Quantile range used to calculate scale_.

copyboolean, optional, default=True

Whether a forced copy will be triggered. If copy=False, a copy might be triggered by a conversion.

Attributes:

center_array of floats

The median value for each feature in the training set.

scale_array of floats

The (scaled) interquartile range for each feature in the training set.

Methods

fit(X[, y])	Compute the median and quantiles to be used for scaling.
inverse_transform(X)	Scale back the data to the original representation
transform(X)	Center and scale the data.

See also

robust_scale

Equivalent function without the estimator API.

cuml.decomposition.PCA

Further removes the linear correlation across features with whiten=True.

Examples

>>> from cuml.preprocessing import RobustScaler
>>> import cupy as cp
>>> X = [[ 1., -2.,  2.],
...      [ -2.,  1.,  3.],
...      [ 4.,  1., -2.]]
>>> X = cp.array(X)
>>> transformer = RobustScaler().fit(X)
>>> transformer
RobustScaler()
>>> transformer.transform(X)
array([[ 0. , -2. ,  0. ],
       [-1. ,  0. ,  0.4],
       [ 1. ,  0. , -1.6]])

fit(X, y=None) → RobustScaler

Compute the median and quantiles to be used for scaling.

Parameters:

X{array-like, CSC matrix}, shape [n_samples, n_features]

The data used to compute the median and quantiles used for later scaling along the features axis.

inverse_transform(X) → SparseCumlArray

Scale back the data to the original representation

Parameters:

X{array-like, sparse matrix}

The data used to scale along the specified axis.

transform(X) → SparseCumlArray

Center and scale the data.

Parameters:

X{array-like, sparse matrix}

The data used to scale along the specified axis.