src.costs.auxi

contained non linear functions (and there derivative) to impose constraints

Module Contents

Functions

f_non_linear(d_min_hard, d_min_soft, d_min_penalization, x)	blow up before d_min_hard and zero after d_min_soft
grad_f_non_linear(d_min_hard, d_min_soft, d_min_penalization, x)	gradient of :func: f_non_linear
f_e(c0, c1, x)	blow up after c0 and zero before c1
grad_f_e(c0, c1, x)	gradient of :func: f_e

src.costs.auxi.f_non_linear(d_min_hard, d_min_soft, d_min_penalization, x)

blow up before d_min_hard and zero after d_min_soft

Parameters

• d_min_hard (float) – cost is inf for x smaller than this value

• d_min_soft (float) – cost is 0 for x bigger than this value

• d_min_penalization (float) – a multiplicative constant in the cost

• x (float) – the argument

Return type

float

src.costs.auxi.grad_f_non_linear(d_min_hard, d_min_soft, d_min_penalization, x)

gradient of :func: f_non_linear

src.costs.auxi.f_e(c0, c1, x)

blow up after c0 and zero before c1

Parameters

• c0 (float) – cost is 0 for x smaller than this value

• c1 (float) – cost is inf for x bigger than this value

• x (float) – the argument

Return type

float

src.costs.auxi.grad_f_e(c0, c1, x)

gradient of :func: f_e