Lines Matching +full:has +full:- +full:flag

2 # Use of this source code is governed by a BSD-style license that can be
14 from flags import Flag
27       first_flag: The first gene (Flag) to cross over with.
28       second_flag: The second gene (Flag) to cross over with.
31       A Flag that can be considered appropriately randomly blended between the
32       first and second input flag.
42       specs: A list of spec from which the flag set is created.
43       flag_set: The current flag set being mutated
47       A Genetic Task constructed by randomly mutating the input flag set.
53         # Randomly choose whether this flag should be mutated.
57         # If the flag is not already in the flag set, it is added.
59             results_flags.append(Flag(spec))
62         # If the flag is already in the flag set, it is mutated.
65         # The value of a numeric flag will be changed, and a boolean flag will be
72         # Randomly select a nearby value of the current value of the flag.
77         rand_arr.append(value - 1)
80         # If the value is smaller than the start of the spec, this flag will be
82         if value != int(numeric_flag_match.group("start")) - 1:
83             results_flags.append(Flag(spec, value))
97           specs: A list of spec from which the flag set is created.
99             random version of the same flag, instead of one from either of the
106         # Get the flag dictionary.
135         # flag from either parent.
136         for flag in common_flags:
138                 CrossoverWith(father_flags[flag], mother_flags[flag])
151     # The value checks whether the algorithm has converged and arrives at a fixed
159     # The value checks whether the algorithm has converged and arrives at a fixed
177           stop_threshold: The number of generations, upon which no performance has
180           num_trials: The number of trials, upon which new task has been tried to
208         """True if this generation has improvement upon its parent generation."""
217         # Found out whether a task has improvement upon the best task in the
222         # At least one task has improvement.