Mixture design is used for formulation optimization (e.g. alloys, coatings, food recipes), where the proportions of components sum to 1 (100%). Supports simplex lattice and simplex centroid designs.
D-optimal design is used for non-standard experimental regions (factors are constrained and a full factorial is not feasible). It selects candidate points that maximize the determinant of the information matrix.
Factor constraints (optional, one per line: factor# low high)