1Blurb:: 2Initial step size for derivative-free optimizers 3 4Description:: 5The \c initial_delta keyword defines the size of the first search step 6in derivative-free optimization methods, specifically \c 7asynch_pattern_search, \c coliny_cobyla, \c coliny_pattern_search, \c 8coliny_solis_wets, and \c mesh_adaptive_search. It is applied in an 9absolute sense to all search directions. 10 11<b> Default Behavior </b> 12 13The default value is 1.0. 14 15<b> Usage Tips </b> 16 17It is recommended that \c initial_delta be the approximate distance 18from the initial point to the solution. If this distance is unknown, 19it is advisable to err on the side of choosing an \c initial_delta 20that is too large or to not specify it. Relative application of \c 21initial_delta is not available unless the user scales the problem 22accordingly. 23 24Topics:: 25 26Examples:: 27 28Three example method input blocks appear below. 29 30For \c asynch_pattern_search: 31\verbatim 32method 33 asynch_pattern_search 34 initial_delta = .5 35 contraction_factor = 0.25 36 merit_function merit1_smooth 37 smoothing_factor = 1.0 38 constraint_tolerance = 1.e-6 39\endverbatim 40 41For \c coliny_pattern_search: 42\verbatim 43method 44 coliny_pattern_search 45 initial_delta = .2 46 variable_tolerance = 1.e-4 47 max_iterations 100 48 solution_accuracy = 1.e-6 49 seed = 1234 50 max_function_evaluations = 1000 51\endverbatim 52 53For \c mesh_adaptive_search 54\verbatim 55method 56 mesh_adaptive_search 57 initial_delta = 2.0 58 seed = 1234 59\endverbatim 60 61Theory:: 62Faq:: 63See_Also:: 64