An acausal modeling framework for automatically parallelized scientific machine learning (SciML) in Julia. A computer algebra system for integrated symbolics for physics-informed machine learning and automated transformations of differential equations
conservative=true limits tearing to only solve for trivial linear systems where the coefficient has the absolute value of 1. This is useful for debugging numerical stability issues after tearing. By default, we set conservative=false.
conservative=truelimits tearing to only solve for trivial linear systems where the coefficient has the absolute value of 1. This is useful for debugging numerical stability issues after tearing. By default, we setconservative=false.