🐛 Add Warning for VahadaneExtractor Algorithm Instability

tiatoolbox/tools/stainextract.py

@@ -5,6 +5,7 @@
 import numpy as np
 from sklearn.decomposition import DictionaryLearning
 
+from tiatoolbox import logger
 from tiatoolbox.utils.misc import get_luminosity_tissue_mask
 from tiatoolbox.utils.transforms import rgb2od
 
@@ -238,6 +239,12 @@ class VahadaneExtractor:
  This class contains code inspired by StainTools
  [https:/Peter554/StainTools] written by Peter Byfield.
 
+ .. warning::
+ Vahadane stain extraction/normalization algorithms are unstable
+ after the update to `dictionary learning` algorithm in
+ scikit-learn > v0.23.0 (see issue #382). Please be advised and
+ consider using other stain extraction (normalization) algorithms.
+
  Args:
  luminosity_threshold (float):
  Threshold used for tissue area selection.
@@ -259,6 +266,14 @@ def __init__(
  regularizer: float = 0.1,
  ) -> None:
  """Initialize :class:`VahadaneExtractor`."""
+ # Issue a warning about the algorithm's stability
+ logger.warning(
+ "Vahadane stain extraction/normalization algorithms are unstable "
+ "after the update to `dictionary learning` algorithm in "
+ "scikit-learn > v0.23.0 (see issue #382). Please be advised and "
+ "consider using other stain extraction (normalization) algorithms.",
+ stacklevel=2,
+ )
  self.__luminosity_threshold = luminosity_threshold
  self.__regularizer = regularizer
 
@@ -267,7 +282,7 @@ def get_stain_matrix(self: VahadaneExtractor, img: np.ndarray) -> np.ndarray:
 
  Args:
  img (:class:`numpy.ndarray`):
- Input image used for stain matrix estimation
+ Input image used for stain matrix estimation.
 
  Returns:
  :class:`numpy.ndarray`: