Partial Least Squares Python Packages

process-improve

Python toolkit for analysis of industrial process data; multivariate analysis, designed experiments, process monitoring.

ikpls

Fast CPU and GPU Python implementations of Improved Kernel Partial Least Squares (PLS) by Dayal and MacGregor (1997) and Fast Partition-Based Cross-Validation With Centering and Scaling for XTX and XTY by Engstrøm and Jensen (2025). This package also includes options to use sample weights for PLS modeling.

cvmatrix

Computation of training set (X^T * W * X) and (X^T * W * Y) or (X^T * X) and (X^T * Y) in a cross-validation setting using the fast algorithms by Engstrøm and Jensen (2025).

scikit-opls

Orthogonal PLS (OPLS / OPLS-DA) with a scikit-learn-compatible API