Source code for chemotools.feature_selection._index_selector
"""
The :mod:`chemotools.feature_selection._index_selector` module implements the IndexSelector
to select specific features from spectral data based on indices or wavenumbers.
"""
# Author: Pau Cabaneros
# License: MIT
from typing import Optional, Union
import numpy as np
from sklearn.base import BaseEstimator
from sklearn.feature_selection._base import SelectorMixin
from sklearn.utils.validation import check_is_fitted, validate_data
[docs]
class IndexSelector(SelectorMixin, BaseEstimator):
"""
A transformer that Selects the spectral data to a specified array of features. This
array can be continuous or discontinuous. The array of features is specified by:
- by the indices of the wavenumbers to select,
- by the wavenumbers to select, the wavenumbers must be provided to the transformer
when it is initialised. If the wavenumbers are not provided, the indices will be
used instead. The wavenumbers must be provided in ascending order.
Parameters
----------
features : narray-like, optional, default=None
The index of the features to select. Default is None.
wavenumbers : array-like, optional, default=None
The wavenumbers of the input data. If not provided, the indices will be used
instead. Default is None. If provided, the wavenumbers must be provided in
ascending order.
Attributes
----------
features_index_ : int
The index of the features to select.
Examples
--------
>>> import numpy as np
>>> from chemotools.feature_selection import IndexSelector
>>> from chemotools.datasets import load_fermentation_train
>>> # Load sample data
>>> X, _ = load_fermentation_train()
>>> # Get wavenumbers as numpy array
>>> wavenumbers = X.columns.values
array([ 428., 429., 431., ..., 1830., 1831., 1833.], shape=(1047,))
>>> # Define features to select
>>> range_1 = np.arange(428, 551, 1)
>>> range_2 = np.arange(875, 1001, 1)
>>> features = np.concatenate((range_1, range_2))
>>> # Instantiate the transformer
>>> selector = IndexSelector(features=features, wavenumbers=wavenumbers)
IndexSelector()
>>> selector.fit(X)
>>> # Transform the data
>>> X_selected = selector.transform(X)
>>> X_selected.shape
(21, 183)
"""
def __init__(
self,
features: Optional[np.ndarray] = None,
wavenumbers: Optional[np.ndarray] = None,
):
self.features = features
self.wavenumbers = wavenumbers
[docs]
def fit(self, X: np.ndarray, y=None) -> "IndexSelector":
"""
Fit the transformer to the input data.
Parameters
----------
X : array-like of shape (n_samples, n_features)
The input data to fit the transformer to.
y : None
Ignored to align with API.
Returns
-------
self : IndexSelector
The fitted transformer.
"""
# validate that X is a 2D array and has only finite values
X = validate_data(
self, X, y="no_validation", ensure_2d=True, reset=True, dtype=np.float64
)
# Set the fitted attribute to True
self._is_fitted = True
# Set the start and end indices
if self.features is None:
self.features_index_ = self.features
return self
elif self.wavenumbers is None:
self.features_index_ = self.features
return self
else:
self.features_index_ = self._find_indices(self.features)
return self
def _get_support_mask(self):
"""
Get the boolean mask indicating which features are selected.
Returns
-------
mask : ndarray of shape (n_features_in_,)
The mask indicating the selected features.
"""
# Check that the estimator is fitted
check_is_fitted(self)
# Create the mask
mask = np.zeros(self.n_features_in_, dtype=bool)
mask[self.features_index_] = True
return mask
def _find_index(self, target: Union[float, int]) -> int:
if self.wavenumbers is None:
return int(target)
wavenumbers = np.array(self.wavenumbers)
return int(np.argmin(np.abs(wavenumbers - target)))
def _find_indices(self, features: np.ndarray) -> np.ndarray:
return np.array([self._find_index(feature) for feature in features])
