scipy.optimize.SR1¶

class scipy.optimize.SR1(min_denominator=1e-08, init_scale='auto')[source]¶

Symmetric-rank-1 Hessian update strategy.

Parameters:

Parameters:	min_denominator : float This number, scaled by a normalization factor, defines the minimum denominator magnitude allowed in the update. When the condition is violated we skip the update. By default uses `1e-8`. init_scale : {float, ‘auto’}, optional Matrix scale at first iteration. At the first iteration the Hessian matrix or its inverse will be initialized with `init_scale*np.eye(n)`, where `n` is the problem dimension. Set it to ‘auto’ in order to use an automatic heuristic for choosing the initial scale. The heuristic is described in [1], p.143. By default uses ‘auto’.

min_denominator : float: This number, scaled by a normalization factor, defines the minimum denominator magnitude allowed in the update. When the condition is violated we skip the update. By default uses 1e-8.
init_scale : {float, ‘auto’}, optional: Matrix scale at first iteration. At the first iteration the Hessian matrix or its inverse will be initialized with init_scale*np.eye(n), where n is the problem dimension. Set it to ‘auto’ in order to use an automatic heuristic for choosing the initial scale. The heuristic is described in [1], p.143. By default uses ‘auto’.

Notes

The update is based on the description in [1], p.144-146.

References

[1]	(1, 2, 3) Nocedal, Jorge, and Stephen J. Wright. “Numerical optimization” Second Edition (2006).

Methods

`dot`(p)	Compute the product of the internal matrix with the given vector.
`get_matrix`()	Return the current internal matrix.
`initialize`(n, approx_type)	Initialize internal matrix.
`update`(delta_x, delta_grad)	Update internal matrix.