pykoop is a Koopman operator identification library written in Python, compatible with scikit-learn. It allows users to specify Koopman lifting functions and regressors to learn linear models of dynamical systems in a lifted space. The library emphasizes modular lifting function construction and leverages scikit-learn's infrastructure for tasks like cross-validation.

Key Features:

• Modular design for customizable lifting functions.
• Integration with scikit-learn for machine learning workflows.
• Support for control inputs and multi-episode datasets.