Optimization of quantum Monte Carlo wave functions using analytical energyderivatives

An algorithm is proposed to optimize quantum Monte Carlo (QMC) wave functio ns based on Newton's method and analytical computation of the first and sec ond derivatives of the variational energy. This direct application of the v ariational principle yields significantly lower energy than variance minimi zation methods when applied to the same trial wave function. Quadratic conv ergence to the local minimum of the variational parameters is achieved. A g eneral theorem is presented, which substantially simplifies the analytic ex pressions of derivatives in the case of wave function optimization. To demo nstrate the method, the ground-state energies of the first-row elements are calculated. (C) 2000 American Institute of Physics. [S0021-9606(00)30605-5 ].