IntrinsicReactionCoordinate

class optking.IRCfollowing.IntrinsicReactionCoordinate(molsys, history, params)

Bases: OptimizationInterface

Methods Summary

add_converged_point(fq, energy)
calc_arc_dist_step()	Let q0 be last rxnpath point and q1 be new rxnpath point.
calc_lagrangian(Lambda, HMEigValues, ...)	Calculates and returns value of Lagrangian function given multiplier Lambda.
calc_lagrangian_derivs(Lambda, HMEigValues, ...)	Calculates and returns value of derivative of Lagrangian function given multiplier Lambda.
calc_line_dist_step()	mass-weighted distance from previous rxnpath point to new one
compute_pivot_and_guess_points(v, fq[, ...])	Takes a half step along v to the 'pivot point', then an additional half step as first guess in constrained opt.
converged(dq, fq, step_number[, str_mode])
dq_irc(f_q, H_q)	Before dq_irc is called, the geometry must be updated to the guess point Returns Dq from qk+1 to gprime.
from_dict(d, molsys, history, params)
requires()
take_step([fq, H, energy, return_str])	Method skeleton (for example see OptimizationAlgorithm) 1.
to_dict()

Methods Documentation

add_converged_point(fq, energy)

calc_arc_dist_step()

Let q0 be last rxnpath point and q1 be new rxnpath point. q* is the pivot point (1/2)s from each of these. Returns the length of circular arc connecting q0 and q1, whose center is equidistant from q0 and q1, and for which line segments from q* to q0 and from q* to q1 are perpendicular to segments from the center to q0 and q1.

calc_lagrangian(Lambda, HMEigValues, HMEigVects, g_M, p_M)

Calculates and returns value of Lagrangian function given multiplier Lambda.

calc_lagrangian_derivs(Lambda, HMEigValues, HMEigVects, g_M, p_M)

Calculates and returns value of derivative of Lagrangian function given multiplier Lambda.

calc_line_dist_step()

mass-weighted distance from previous rxnpath point to new one

compute_pivot_and_guess_points(v, fq, return_str=False)

Takes a half step along v to the ‘pivot point’, then an additional half step as first guess in constrained opt.

Parameters

v (ndarray) – initial vector to step along. Hessian eigenvector for first step. Gradient at subsequent steps

converged(dq, fq, step_number, str_mode='', **kwargs)

dq_irc(f_q, H_q)

Before dq_irc is called, the geometry must be updated to the guess point Returns Dq from qk+1 to gprime. TODO: What is dqGuess for? Remove it?

classmethod from_dict(d, molsys, history, params)

requires()

take_step(fq=None, H=None, energy=None, return_str=False, **kwargs)

Method skeleton (for example see OptimizationAlgorithm) 1. Choose what kind of step should take place next 2. take step 3. displace molsys 4. update history (trim history as applicable) 5. return step taken

to_dict()