PyDeepCausalML 0.2.0

Major expansion of causal effect, discovery, generative, and time-series models, with CPU / GPU / MPS device support.

Added

• effect: GANITE, CEVAE
• generative: CausalGAN, CausalDiscrepancyVAE, iVAE, CausalVAE, DSCM
• discovery: nonlinear NOTEARS, DAG-GNN, DAGMA, and the causal_structure_ml factory
• timeseries: neural Granger, attention, RNN, GNN, and counterfactual model families
• device utilities with automatic CUDA, MPS, then CPU selection

Changed

• Estimators accept device=None (auto) or an explicit cpu, cuda, or mps device
• scipy is now a core dependency

Assets are the 0.2.0 wheel and source distribution.