Acceptable objective change tolerance

Type

Floating point number

Range

[0,1e+019]

Default

1e+019

This option specifies the "acceptance" stopping criterion based on objective function change. If the relative change of the objective function (scaled by Max(1,|f(x)|)) is less than this value, this part of the acceptable tolerance termination is satisfied; see also the option Acceptable Relative Convergence Tolerance . This is useful for the quasi-Newton option, which has trouble to bring down the dual infeasibility.

Learn more about

• Acceptable relative convergence tolerance