Preprocessing¶

The sklearn_xarray.preprocessing module contains various preprocessing methods that work on xarray DataArrays and Datasets.

class BaseTransformer[source]¶

Base class for transformers.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

fit(self, X, y=None, **fit_params)[source]¶

Fit estimator to data.

Parameters

Xxarray DataArray or Dataset: Training set.
yxarray DataArray or Dataset: Target values.

Returns

self:: The estimator itself.

inverse_transform(self, X)[source]¶

Reverse the transformation.

Parameters

Xxarray DataArray or Dataset: The input data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

transform(self, X)[source]¶

Transform input data.

Parameters

Xxarray DataArray or Dataset: The input data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

class Concatenator(dim='feature', new_dim=None, variables=None, new_var='Feature', new_index_func=None, return_array=False, groupby=None, group_dim='sample')[source]¶

Concatenate variables along a dimension.

Parameters

dimstr: Name of the dimension along which to concatenate the Dataset.
new_dimstr: New name of the dimension, if desired.
variableslist or tuple: Names of the variables to concatenate, default all.
new_var :: Name of the new variable created by the concatenation.
new_index_funcfunction: A function that takes the length of the concatenated dimension as a parameter and returns a vector of this length to be used as the index for that dimension.
return_array: bool: Whether to return a DataArray when a Dataset was passed.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

fit(self, X, y=None, **fit_params)[source]¶

Fit estimator to data.

Parameters

Xxarray DataArray or Dataset: Training set.
yxarray DataArray or Dataset: Target values.

Returns

self:: The estimator itself.

class Featurizer(sample_dim='sample', feature_dim='feature', var_name='Features', order=None, return_array=False, groupby=None, group_dim='sample')[source]¶

Stack all dimensions and variables except for sample dimension.

Parameters

sample_dimstr: Name of the sample dimension.
feature_dimstr: Name of the feature dimension.
var_namestr: Name of the new variable (for Datasets).
orderlist or tuple: Order of dimension stacking.
return_array: bool: Whether to return a DataArray when a Dataset was passed.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

class Reducer(dim='feature', func=<function norm>, groupby=None, group_dim='sample')[source]¶

Reduce data along some dimension.

Parameters

dimstr: Name of the dimension.
funcfunction: Reduction function.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

class Resampler(freq=None, dim='sample', groupby=None, group_dim='sample')[source]¶

Resample along some dimension.

Parameters

freqstr: Frequency after resampling.
dimstr: Name of the dimension along which to resample.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit the estimator.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

fit(self, X, y=None, **fit_params)[source]¶

Fit the estimator.

Parameters

Xxarray DataArray or Dataset: The input data.
yNone: For compatibility.

Returns

self :: The estimator itself.

class Sanitizer(dim='sample', groupby=None, group_dim='sample')[source]¶

Remove elements containing NaNs.

Parameters

dimstr: Name of the sample dimension.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

class Segmenter(dim='sample', new_dim=None, new_len=None, step=None, axis=None, reduce_index='subsample', new_index_func=<built-in function arange>, keep_coords_as=None, groupby=None, group_dim='sample', return_view=False)[source]¶

Split into segments along some dimension.

Parameters

dimstr

Name of the dimension along which to split.

new_dimstr

Name of the newly added dimension.

new_lenint

Length of the newly added dimension.

step: int

Number of values between the start of a segment and the next one.

axisint

Axis position where new dimension is to be inserted. If None, the dimension will be inserted at the end.

reduce_indexstr

How to reduce the index of the split dimension.

'head' : Take the first n values where n is the length of the dimension after segmenting.
'subsample' : Take the values corresponding to the first element of every segment.

new_index_funccallable

A function that takes new_len as a parameter and returns a vector of length new_len to be used as the indices for the new dimension.

keep_coords_asstr or None

If set, the coordinate of the split dimension will be kept as a separate coordinate with this name. This allows inverse_transform to reconstruct the original coordinate.

return_viewbool, default False

If true, return a view instead of a copy of the segmented array.

groupbystr or list, optional

Name of coordinate or list of coordinates by which the groups are determined.

group_dimstr, optional

Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

class Selector(dim='sample', coord=None, groupby=None, group_dim='sample')[source]¶

Selects a subset of the samples.

Parameters

dimstr: Name of the sample dimension.
coordstr: The name of the coordinate that acts as the selector.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

class Splitter(dim='sample', new_dim=None, new_len=None, axis=None, reduce_index='subsample', new_index_func=<built-in function arange>, keep_coords_as=None, groupby=None, group_dim='sample')[source]¶

Split along some dimension.

Parameters

dimstr

Name of the dimension along which to split.

new_dimstr

Name of the newly added dimension.

new_lenint

Length of the newly added dimension.

axisint

Axis position where new dimension is to be inserted. If None, the dimension will be inserted at the end.

reduce_indexstr

How to reduce the index of the split dimension.

'head' : Take the first n values where n is the length of the dimension after splitting.
'subsample' : Take every new_len th value.

new_index_funccallable

A function that takes new_len as a parameter and returns a vector of length new_len to be used as the indices for the new dimension.

keep_coords_asstr or None

If set, the coordinate of the split dimension will be kept as a separate coordinate with this name. This allows inverse_transform to reconstruct the original coordinate.

groupbystr or list, optional

Name of coordinate or list of coordinates by which the groups are determined.

group_dimstr, optional

Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit estimator to data.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

class Transposer(order=None, groupby=None, group_dim='sample')[source]¶

Reorder data dimensions.

Parameters

orderlist or tuple: The new order of the dimensions.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Methods

`fit`(self, X[, y])	Fit the estimator.
`fit_transform`(self, X[, y])	Fit to data, then transform it.
`get_params`(self[, deep])	Get parameters for this estimator.
`inverse_transform`(self, X)	Reverse the transformation.
`set_params`(self, **params)	Set the parameters of this estimator.
`transform`(self, X)	Transform input data.

fit(self, X, y=None, **fit_params)[source]¶

Fit the estimator.

Parameters

Xxarray DataArray or Dataset: The input data.
yNone: For compatibility.

Returns

self :: The estimator itself.

concatenate(X, return_estimator=False, **fit_params)[source]¶

Concatenates variables along a dimension.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

featurize(X, return_estimator=False, **fit_params)[source]¶

Stacks all dimensions and variables except for sample dimension.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

preprocess(X, function, groupby=None, group_dim='sample', **fit_params)[source]¶

Wraps preprocessing functions from sklearn for use with xarray types.

Parameters

Xxarray DataArray or Dataset: The input data.
functioncallable: The function to apply to the data. Note that this function cannot change the shape of the data.
groupbystr or list, optional: Name of coordinate or list of coordinates by which the groups are determined.
group_dimstr, optional: Name of dimension along which the groups are indexed.

Returns

Xtxarray DataArray or Dataset: The transformed data.

reduce(X, return_estimator=False, **fit_params)[source]¶

Reduces data along some dimension.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

resample(X, return_estimator=False, **fit_params)[source]¶

Resamples along some dimension.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

sanitize(X, return_estimator=False, **fit_params)[source]¶

Removes elements containing NaNs.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

segment(X, return_estimator=False, **fit_params)[source]¶

Segments X along some dimension.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

select(X, return_estimator=False, **fit_params)[source]¶

Selects a subset of the samples.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

split(X, return_estimator=False, **fit_params)[source]¶

Splits X along some dimension.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.

transpose(X, return_estimator=False, **fit_params)[source]¶

Reorders data dimensions.

Parameters

Xxarray DataArray or Dataset: The input data.
return_estimatorbool: Whether to return the fitted estimator along with the transformed data.

Returns

Xtxarray DataArray or Dataset: The transformed data.