I just want to know whether I can use the Lagrange multipliers method to solve a problems with complex-valued vector variables? Should I redefine constraints based on real-valued vector variables?, or I can use it with complex-valued vector variables?