VERSION 1.2.1
------------------------
   o Replaced geigen dependency with iTensor (>= 1.0.6) for SUMCOR-based CCA
   o geigen package is scheduled for deprecation; iTensor::geigen is the replacement
   o Updated Suggests: removed geigen, added iTensor (>= 1.0.6)

   VERSION 1.2.0
------------------------
   o Added sparse matrix support (Matrix::dgCMatrix) for guidedPLS()
   o X1, X2 can now be sparse matrices; avoids dense materialization of centered X
   o Internally uses the identity X_c Y = X Y - mu (1^T Y) to keep X sparse
   o Added Matrix to Imports
   o Particularly effective for scATAC-seq peak matrices with low density

   VERSION 1.1.0
------------------------
   o Added guidedPCA() function for guided principal component analysis
   o guidedPCA() automatically handles mixed data types (continuous, categorical, logical)
   o Added automatic dummy variable encoding for categorical variables
   o Added feature contribution analysis to identify important features per component
   o Added variance explained calculation for each component
   o Added print() and summary() methods for guidedPCA objects
   o Added comprehensive unit tests for guidedPCA()
   o Updated package documentation with detailed algorithm descriptions
   o Added 'stats' to Imports for better dependency management

   VERSION 1.0.0
------------------------
   o Added cortest in PLSSVD() and sPLSDA()

   VERSION 0.99.0
------------------------
   o Package released