After diagonalising the small-oscillation energy of a triatomic molecule

PHYSICS · Multi-part · Fermi University Physics · Computational Physics · Scientific programming for physicists · Scientific programming for physicists

Scientific Computing for Physicists, Ch Minimisation with guarantees, with constraints and with randomness

Problem

After diagonalising the small-oscillation energy of a triatomic molecule, two of its normal coordinates (x,y)(x,y) contribute a quadratic energy f(x,y)=450x2+0.5y2 (arbitrary energy units).f(x,y) = 450x^2 + 0.5y^2 \ \text{(arbitrary energy units)}. ((a)) Predict, from the condition number of the Hessian, the number of steepest-descent steps (with exact line search) needed to reduce f−f∗f-f^* by four orders of magnitude from the worst starting direction, a count no other start can exceed. ((b)) Predict, without doing the arithmetic of a full run, how many iterations Newton's method needs to reach the minimum exactly from any starting point, and explain why the answer does not depend on κ\kappa.

Browse the problem library