mono.html 34 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511
  1. <!DOCTYPE html>
  2. <!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
  3. <!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
  4. <head>
  5. <meta charset="utf-8">
  6. <meta name="viewport" content="width=device-width, initial-scale=1.0">
  7. <title>macop.algorithms.mono &mdash; macop v1.0.5 documentation</title>
  8. <script type="text/javascript" src="../../../_static/js/modernizr.min.js"></script>
  9. <script type="text/javascript" id="documentation_options" data-url_root="../../../" src="../../../_static/documentation_options.js"></script>
  10. <script type="text/javascript" src="../../../_static/jquery.js"></script>
  11. <script type="text/javascript" src="../../../_static/underscore.js"></script>
  12. <script type="text/javascript" src="../../../_static/doctools.js"></script>
  13. <script type="text/javascript" src="../../../_static/language_data.js"></script>
  14. <script type="text/javascript" src="../../../_static/js/theme.js"></script>
  15. <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
  16. <link rel="stylesheet" href="../../../_static/pygments.css" type="text/css" />
  17. <link rel="stylesheet" href="../../../_static/css/custom.css" type="text/css" />
  18. <link rel="index" title="Index" href="../../../genindex.html" />
  19. <link rel="search" title="Search" href="../../../search.html" />
  20. </head>
  21. <body class="wy-body-for-nav">
  22. <div class="wy-grid-for-nav">
  23. <nav data-toggle="wy-nav-shift" class="wy-nav-side">
  24. <div class="wy-side-scroll">
  25. <div class="wy-side-nav-search" >
  26. <a href="../../../index.html" class="icon icon-home"> macop
  27. </a>
  28. <div class="version">
  29. 1.0.5
  30. </div>
  31. <div role="search">
  32. <form id="rtd-search-form" class="wy-form" action="../../../search.html" method="get">
  33. <input type="text" name="q" placeholder="Search docs" />
  34. <input type="hidden" name="check_keywords" value="yes" />
  35. <input type="hidden" name="area" value="default" />
  36. </form>
  37. </div>
  38. </div>
  39. <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
  40. <p class="caption"><span class="caption-text">Contents:</span></p>
  41. <ul>
  42. <li class="toctree-l1"><a class="reference internal" href="../../../description.html">Description</a></li>
  43. <li class="toctree-l1"><a class="reference internal" href="../../../api.html">API</a></li>
  44. <li class="toctree-l1"><a class="reference internal" href="../../../documentations/index.html">Documentation</a></li>
  45. <li class="toctree-l1"><a class="reference internal" href="../../../examples.html">Some examples</a></li>
  46. <li class="toctree-l1"><a class="reference internal" href="../../../contributing.html">Contributing</a></li>
  47. </ul>
  48. </div>
  49. </div>
  50. </nav>
  51. <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">
  52. <nav class="wy-nav-top" aria-label="top navigation">
  53. <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
  54. <a href="../../../index.html">macop</a>
  55. </nav>
  56. <div class="wy-nav-content">
  57. <div class="rst-content">
  58. <div role="navigation" aria-label="breadcrumbs navigation">
  59. <ul class="wy-breadcrumbs">
  60. <li><a href="../../../index.html">Docs</a> &raquo;</li>
  61. <li><a href="../../index.html">Module code</a> &raquo;</li>
  62. <li>macop.algorithms.mono</li>
  63. <li class="wy-breadcrumbs-aside">
  64. </li>
  65. </ul>
  66. <hr/>
  67. </div>
  68. <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
  69. <div itemprop="articleBody">
  70. <h1>Source code for macop.algorithms.mono</h1><div class="highlight"><pre>
  71. <span></span><span class="sd">&quot;&quot;&quot;Mono-objective available algorithms</span>
  72. <span class="sd">&quot;&quot;&quot;</span>
  73. <span class="c1"># main imports</span>
  74. <span class="kn">import</span> <span class="nn">logging</span>
  75. <span class="c1"># module imports</span>
  76. <span class="kn">from</span> <span class="nn">.base</span> <span class="kn">import</span> <span class="n">Algorithm</span>
  77. <div class="viewcode-block" id="HillClimberFirstImprovment"><a class="viewcode-back" href="../../../macop/macop.algorithms.mono.html#macop.algorithms.mono.HillClimberFirstImprovment">[docs]</a><span class="k">class</span> <span class="nc">HillClimberFirstImprovment</span><span class="p">(</span><span class="n">Algorithm</span><span class="p">):</span>
  78. <span class="sd">&quot;&quot;&quot;Hill Climber First Improvment used as quick exploration optimisation algorithm</span>
  79. <span class="sd"> - First, this algorithm do a neighborhood exploration of a new generated solution (by doing operation on the current solution obtained) in order to find a better solution from the neighborhood space.</span>
  80. <span class="sd"> - Then replace the current solution by the first one from the neighbordhood space which is better than the current solution.</span>
  81. <span class="sd"> - And do these steps until a number of evaluation (stopping criterion) is reached.</span>
  82. <span class="sd"> Attributes:</span>
  83. <span class="sd"> initalizer: {function} -- basic function strategy to initialize solution</span>
  84. <span class="sd"> evaluator: {function} -- basic function in order to obtained fitness (mono or multiple objectives)</span>
  85. <span class="sd"> operators: {[Operator]} -- list of operator to use when launching algorithm</span>
  86. <span class="sd"> policy: {Policy} -- Policy class implementation strategy to select operators</span>
  87. <span class="sd"> validator: {function} -- basic function to check if solution is valid or not under some constraints</span>
  88. <span class="sd"> maximise: {bool} -- specify kind of optimisation problem </span>
  89. <span class="sd"> currentSolution: {Solution} -- current solution managed for current evaluation</span>
  90. <span class="sd"> bestSolution: {Solution} -- best solution found so far during running algorithm</span>
  91. <span class="sd"> callbacks: {[Callback]} -- list of Callback class implementation to do some instructions every number of evaluations and `load` when initializing algorithm</span>
  92. <span class="sd"> </span>
  93. <span class="sd"> Example:</span>
  94. <span class="sd"> &gt;&gt;&gt; import random</span>
  95. <span class="sd"> &gt;&gt;&gt; # operators import</span>
  96. <span class="sd"> &gt;&gt;&gt; from macop.operators.discrete.crossovers import SimpleCrossover</span>
  97. <span class="sd"> &gt;&gt;&gt; from macop.operators.discrete.mutators import SimpleMutation</span>
  98. <span class="sd"> &gt;&gt;&gt; # policy import</span>
  99. <span class="sd"> &gt;&gt;&gt; from macop.policies.classicals import RandomPolicy</span>
  100. <span class="sd"> &gt;&gt;&gt; # solution and algorithm</span>
  101. <span class="sd"> &gt;&gt;&gt; from macop.solutions.discrete import BinarySolution</span>
  102. <span class="sd"> &gt;&gt;&gt; from macop.algorithms.mono import HillClimberFirstImprovment</span>
  103. <span class="sd"> &gt;&gt;&gt; # evaluator import</span>
  104. <span class="sd"> &gt;&gt;&gt; from macop.evaluators.discrete.mono import KnapsackEvaluator</span>
  105. <span class="sd"> &gt;&gt;&gt; # evaluator initialization (worths objects passed into data)</span>
  106. <span class="sd"> &gt;&gt;&gt; problem_size = 20</span>
  107. <span class="sd"> &gt;&gt;&gt; worths = [ random.randint(0, 20) for i in range(problem_size) ]</span>
  108. <span class="sd"> &gt;&gt;&gt; evaluator = KnapsackEvaluator(data={&#39;worths&#39;: worths})</span>
  109. <span class="sd"> &gt;&gt;&gt; # validator specification (based on weights of each objects)</span>
  110. <span class="sd"> &gt;&gt;&gt; weights = [ random.randint(5, 30) for i in range(problem_size) ]</span>
  111. <span class="sd"> &gt;&gt;&gt; validator = lambda solution: True if sum([weights[i] for i, value in enumerate(solution._data) if value == 1]) &lt; 200 else False</span>
  112. <span class="sd"> &gt;&gt;&gt; # initializer function with lambda function</span>
  113. <span class="sd"> &gt;&gt;&gt; initializer = lambda x=20: BinarySolution.random(x, validator)</span>
  114. <span class="sd"> &gt;&gt;&gt; # operators list with crossover and mutation</span>
  115. <span class="sd"> &gt;&gt;&gt; operators = [SimpleCrossover(), SimpleMutation()]</span>
  116. <span class="sd"> &gt;&gt;&gt; policy = RandomPolicy(operators)</span>
  117. <span class="sd"> &gt;&gt;&gt; algo = HillClimberFirstImprovment(initializer, evaluator, operators, policy, validator, maximise=True, verbose=False)</span>
  118. <span class="sd"> &gt;&gt;&gt; # run the algorithm</span>
  119. <span class="sd"> &gt;&gt;&gt; solution = algo.run(100)</span>
  120. <span class="sd"> &gt;&gt;&gt; solution._score</span>
  121. <span class="sd"> 128</span>
  122. <span class="sd"> &quot;&quot;&quot;</span>
  123. <div class="viewcode-block" id="HillClimberFirstImprovment.run"><a class="viewcode-back" href="../../../macop/macop.algorithms.mono.html#macop.algorithms.mono.HillClimberFirstImprovment.run">[docs]</a> <span class="k">def</span> <span class="nf">run</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">evaluations</span><span class="p">):</span>
  124. <span class="sd">&quot;&quot;&quot;</span>
  125. <span class="sd"> Run the local search algorithm</span>
  126. <span class="sd"> Args:</span>
  127. <span class="sd"> evaluations: {int} -- number of Local search evaluations</span>
  128. <span class="sd"> </span>
  129. <span class="sd"> Returns:</span>
  130. <span class="sd"> {Solution} -- best solution found</span>
  131. <span class="sd"> &quot;&quot;&quot;</span>
  132. <span class="c1"># by default use of mother method to initialize variables</span>
  133. <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="n">run</span><span class="p">(</span><span class="n">evaluations</span><span class="p">)</span>
  134. <span class="c1"># initialize current solution and best solution</span>
  135. <span class="bp">self</span><span class="o">.</span><span class="n">initRun</span><span class="p">()</span>
  136. <span class="n">solutionSize</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_currentSolution</span><span class="o">.</span><span class="n">_size</span>
  137. <span class="c1"># local search algorithm implementation</span>
  138. <span class="k">while</span> <span class="ow">not</span> <span class="bp">self</span><span class="o">.</span><span class="n">stop</span><span class="p">():</span>
  139. <span class="k">for</span> <span class="n">_</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="n">solutionSize</span><span class="p">):</span>
  140. <span class="c1"># update current solution using policy</span>
  141. <span class="n">newSolution</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">update</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">_currentSolution</span><span class="p">)</span>
  142. <span class="c1"># if better solution than currently, replace it and stop current exploration (first improvment)</span>
  143. <span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">isBetter</span><span class="p">(</span><span class="n">newSolution</span><span class="p">):</span>
  144. <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span> <span class="o">=</span> <span class="n">newSolution</span>
  145. <span class="k">break</span>
  146. <span class="c1"># increase number of evaluations</span>
  147. <span class="bp">self</span><span class="o">.</span><span class="n">increaseEvaluation</span><span class="p">()</span>
  148. <span class="bp">self</span><span class="o">.</span><span class="n">progress</span><span class="p">()</span>
  149. <span class="n">logging</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">&quot;---- Current </span><span class="si">{</span><span class="n">newSolution</span><span class="si">}</span><span class="s2"> - SCORE </span><span class="si">{</span><span class="n">newSolution</span><span class="o">.</span><span class="n">fitness</span><span class="p">()</span><span class="si">}</span><span class="s2">&quot;</span><span class="p">)</span>
  150. <span class="c1"># stop algorithm if necessary</span>
  151. <span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">stop</span><span class="p">():</span>
  152. <span class="k">break</span>
  153. <span class="c1"># set new current solution using best solution found in this neighbor search</span>
  154. <span class="bp">self</span><span class="o">.</span><span class="n">_currentSolution</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span>
  155. <span class="n">logging</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">&quot;End of </span><span class="si">{</span><span class="nb">type</span><span class="p">(</span><span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="vm">__name__</span><span class="si">}</span><span class="s2">, best solution found </span><span class="si">{</span><span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span><span class="si">}</span><span class="s2">&quot;</span><span class="p">)</span>
  156. <span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span></div></div>
  157. <div class="viewcode-block" id="HillClimberBestImprovment"><a class="viewcode-back" href="../../../macop/macop.algorithms.mono.html#macop.algorithms.mono.HillClimberBestImprovment">[docs]</a><span class="k">class</span> <span class="nc">HillClimberBestImprovment</span><span class="p">(</span><span class="n">Algorithm</span><span class="p">):</span>
  158. <span class="sd">&quot;&quot;&quot;Hill Climber Best Improvment used as exploitation optimisation algorithm</span>
  159. <span class="sd"> - First, this algorithm do a neighborhood exploration of a new generated solution (by doing operation on the current solution obtained) in order to find the best solution from the neighborhood space.</span>
  160. <span class="sd"> - Then replace the best solution found from the neighbordhood space as current solution to use.</span>
  161. <span class="sd"> - And do these steps until a number of evaluation (stopping criterion) is reached.</span>
  162. <span class="sd"> Attributes:</span>
  163. <span class="sd"> initalizer: {function} -- basic function strategy to initialize solution</span>
  164. <span class="sd"> evaluator: {function} -- basic function in order to obtained fitness (mono or multiple objectives)</span>
  165. <span class="sd"> operators: {[Operator]} -- list of operator to use when launching algorithm</span>
  166. <span class="sd"> policy: {Policy} -- Policy class implementation strategy to select operators</span>
  167. <span class="sd"> validator: {function} -- basic function to check if solution is valid or not under some constraints</span>
  168. <span class="sd"> maximise: {bool} -- specify kind of optimisation problem </span>
  169. <span class="sd"> currentSolution: {Solution} -- current solution managed for current evaluation</span>
  170. <span class="sd"> bestSolution: {Solution} -- best solution found so far during running algorithm</span>
  171. <span class="sd"> callbacks: {[Callback]} -- list of Callback class implementation to do some instructions every number of evaluations and `load` when initializing algorithm</span>
  172. <span class="sd"> </span>
  173. <span class="sd"> Example:</span>
  174. <span class="sd"> &gt;&gt;&gt; import random</span>
  175. <span class="sd"> &gt;&gt;&gt; # operators import</span>
  176. <span class="sd"> &gt;&gt;&gt; from macop.operators.discrete.crossovers import SimpleCrossover</span>
  177. <span class="sd"> &gt;&gt;&gt; from macop.operators.discrete.mutators import SimpleMutation</span>
  178. <span class="sd"> &gt;&gt;&gt; # policy import</span>
  179. <span class="sd"> &gt;&gt;&gt; from macop.policies.classicals import RandomPolicy</span>
  180. <span class="sd"> &gt;&gt;&gt; # solution and algorithm</span>
  181. <span class="sd"> &gt;&gt;&gt; from macop.solutions.discrete import BinarySolution</span>
  182. <span class="sd"> &gt;&gt;&gt; from macop.algorithms.mono import HillClimberBestImprovment</span>
  183. <span class="sd"> &gt;&gt;&gt; # evaluator import</span>
  184. <span class="sd"> &gt;&gt;&gt; from macop.evaluators.discrete.mono import KnapsackEvaluator</span>
  185. <span class="sd"> &gt;&gt;&gt; # evaluator initialization (worths objects passed into data)</span>
  186. <span class="sd"> &gt;&gt;&gt; problem_size = 20</span>
  187. <span class="sd"> &gt;&gt;&gt; worths = [ random.randint(0, 20) for i in range(problem_size) ]</span>
  188. <span class="sd"> &gt;&gt;&gt; evaluator = KnapsackEvaluator(data={&#39;worths&#39;: worths})</span>
  189. <span class="sd"> &gt;&gt;&gt; # validator specification (based on weights of each objects)</span>
  190. <span class="sd"> &gt;&gt;&gt; weights = [ random.randint(5, 30) for i in range(problem_size) ]</span>
  191. <span class="sd"> &gt;&gt;&gt; validator = lambda solution: True if sum([weights[i] for i, value in enumerate(solution._data) if value == 1]) &lt; 200 else False</span>
  192. <span class="sd"> &gt;&gt;&gt; # initializer function with lambda function</span>
  193. <span class="sd"> &gt;&gt;&gt; initializer = lambda x=20: BinarySolution.random(x, validator)</span>
  194. <span class="sd"> &gt;&gt;&gt; # operators list with crossover and mutation</span>
  195. <span class="sd"> &gt;&gt;&gt; operators = [SimpleCrossover(), SimpleMutation()]</span>
  196. <span class="sd"> &gt;&gt;&gt; policy = RandomPolicy(operators)</span>
  197. <span class="sd"> &gt;&gt;&gt; algo = HillClimberBestImprovment(initializer, evaluator, operators, policy, validator, maximise=True, verbose=False)</span>
  198. <span class="sd"> &gt;&gt;&gt; # run the algorithm</span>
  199. <span class="sd"> &gt;&gt;&gt; solution = algo.run(100)</span>
  200. <span class="sd"> &gt;&gt;&gt; solution._score</span>
  201. <span class="sd"> 104</span>
  202. <span class="sd"> &quot;&quot;&quot;</span>
  203. <div class="viewcode-block" id="HillClimberBestImprovment.run"><a class="viewcode-back" href="../../../macop/macop.algorithms.mono.html#macop.algorithms.mono.HillClimberBestImprovment.run">[docs]</a> <span class="k">def</span> <span class="nf">run</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">evaluations</span><span class="p">):</span>
  204. <span class="sd">&quot;&quot;&quot;</span>
  205. <span class="sd"> Run the local search algorithm</span>
  206. <span class="sd"> Args:</span>
  207. <span class="sd"> evaluations: {int} -- number of Local search evaluations</span>
  208. <span class="sd"> </span>
  209. <span class="sd"> Returns:</span>
  210. <span class="sd"> {Solution} -- best solution found</span>
  211. <span class="sd"> &quot;&quot;&quot;</span>
  212. <span class="c1"># by default use of mother method to initialize variables</span>
  213. <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="n">run</span><span class="p">(</span><span class="n">evaluations</span><span class="p">)</span>
  214. <span class="c1"># initialize current solution and best solution</span>
  215. <span class="bp">self</span><span class="o">.</span><span class="n">initRun</span><span class="p">()</span>
  216. <span class="n">solutionSize</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_currentSolution</span><span class="o">.</span><span class="n">_size</span>
  217. <span class="c1"># local search algorithm implementation</span>
  218. <span class="k">while</span> <span class="ow">not</span> <span class="bp">self</span><span class="o">.</span><span class="n">stop</span><span class="p">():</span>
  219. <span class="k">for</span> <span class="n">_</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="n">solutionSize</span><span class="p">):</span>
  220. <span class="c1"># update current solution using policy</span>
  221. <span class="n">newSolution</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">update</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">_currentSolution</span><span class="p">)</span>
  222. <span class="c1"># if better solution than currently, replace it</span>
  223. <span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">isBetter</span><span class="p">(</span><span class="n">newSolution</span><span class="p">):</span>
  224. <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span> <span class="o">=</span> <span class="n">newSolution</span>
  225. <span class="c1"># increase number of evaluations</span>
  226. <span class="bp">self</span><span class="o">.</span><span class="n">increaseEvaluation</span><span class="p">()</span>
  227. <span class="bp">self</span><span class="o">.</span><span class="n">progress</span><span class="p">()</span>
  228. <span class="n">logging</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">&quot;---- Current </span><span class="si">{</span><span class="n">newSolution</span><span class="si">}</span><span class="s2"> - SCORE </span><span class="si">{</span><span class="n">newSolution</span><span class="o">.</span><span class="n">fitness</span><span class="p">()</span><span class="si">}</span><span class="s2">&quot;</span><span class="p">)</span>
  229. <span class="c1"># stop algorithm if necessary</span>
  230. <span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">stop</span><span class="p">():</span>
  231. <span class="k">break</span>
  232. <span class="c1"># set new current solution using best solution found in this neighbor search</span>
  233. <span class="bp">self</span><span class="o">.</span><span class="n">_currentSolution</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span>
  234. <span class="n">logging</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">&quot;End of </span><span class="si">{</span><span class="nb">type</span><span class="p">(</span><span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="vm">__name__</span><span class="si">}</span><span class="s2">, best solution found </span><span class="si">{</span><span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span><span class="si">}</span><span class="s2">&quot;</span><span class="p">)</span>
  235. <span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span></div></div>
  236. <div class="viewcode-block" id="IteratedLocalSearch"><a class="viewcode-back" href="../../../macop/macop.algorithms.mono.html#macop.algorithms.mono.IteratedLocalSearch">[docs]</a><span class="k">class</span> <span class="nc">IteratedLocalSearch</span><span class="p">(</span><span class="n">Algorithm</span><span class="p">):</span>
  237. <span class="sd">&quot;&quot;&quot;Iterated Local Search used to avoid local optima and increave EvE (Exploration vs Exploitation) compromise</span>
  238. <span class="sd"> - A number of evaluations (`ls_evaluations`) is dedicated to local search process, here `HillClimberFirstImprovment` algorithm</span>
  239. <span class="sd"> - Starting with the new generated solution, the local search algorithm will return a new solution</span>
  240. <span class="sd"> - If the obtained solution is better than the best solution known into `IteratedLocalSearch`, then the solution is replaced</span>
  241. <span class="sd"> - Restart this process until stopping critirion (number of expected evaluations)</span>
  242. <span class="sd"> Attributes:</span>
  243. <span class="sd"> initalizer: {function} -- basic function strategy to initialize solution</span>
  244. <span class="sd"> evaluator: {function} -- basic function in order to obtained fitness (mono or multiple objectives)</span>
  245. <span class="sd"> operators: {[Operator]} -- list of operator to use when launching algorithm</span>
  246. <span class="sd"> policy: {Policy} -- Policy class implementation strategy to select operators</span>
  247. <span class="sd"> validator: {function} -- basic function to check if solution is valid or not under some constraints</span>
  248. <span class="sd"> maximise: {bool} -- specify kind of optimisation problem </span>
  249. <span class="sd"> currentSolution: {Solution} -- current solution managed for current evaluation</span>
  250. <span class="sd"> bestSolution: {Solution} -- best solution found so far during running algorithm</span>
  251. <span class="sd"> callbacks: {[Callback]} -- list of Callback class implementation to do some instructions every number of evaluations and `load` when initializing algorithm</span>
  252. <span class="sd"> </span>
  253. <span class="sd"> Example:</span>
  254. <span class="sd"> &gt;&gt;&gt; import random</span>
  255. <span class="sd"> &gt;&gt;&gt; # operators import</span>
  256. <span class="sd"> &gt;&gt;&gt; from macop.operators.discrete.crossovers import SimpleCrossover</span>
  257. <span class="sd"> &gt;&gt;&gt; from macop.operators.discrete.mutators import SimpleMutation</span>
  258. <span class="sd"> &gt;&gt;&gt; # policy import</span>
  259. <span class="sd"> &gt;&gt;&gt; from macop.policies.classicals import RandomPolicy</span>
  260. <span class="sd"> &gt;&gt;&gt; # solution and algorithm</span>
  261. <span class="sd"> &gt;&gt;&gt; from macop.solutions.discrete import BinarySolution</span>
  262. <span class="sd"> &gt;&gt;&gt; from macop.algorithms.mono import IteratedLocalSearch</span>
  263. <span class="sd"> &gt;&gt;&gt; # evaluator import</span>
  264. <span class="sd"> &gt;&gt;&gt; from macop.evaluators.discrete.mono import KnapsackEvaluator</span>
  265. <span class="sd"> &gt;&gt;&gt; # evaluator initialization (worths objects passed into data)</span>
  266. <span class="sd"> &gt;&gt;&gt; problem_size = 20</span>
  267. <span class="sd"> &gt;&gt;&gt; worths = [ random.randint(0, 20) for i in range(problem_size) ]</span>
  268. <span class="sd"> &gt;&gt;&gt; evaluator = KnapsackEvaluator(data={&#39;worths&#39;: worths})</span>
  269. <span class="sd"> &gt;&gt;&gt; # validator specification (based on weights of each objects)</span>
  270. <span class="sd"> &gt;&gt;&gt; weights = [ random.randint(5, 30) for i in range(problem_size) ]</span>
  271. <span class="sd"> &gt;&gt;&gt; validator = lambda solution: True if sum([weights[i] for i, value in enumerate(solution._data) if value == 1]) &lt; 200 else False</span>
  272. <span class="sd"> &gt;&gt;&gt; # initializer function with lambda function</span>
  273. <span class="sd"> &gt;&gt;&gt; initializer = lambda x=20: BinarySolution.random(x, validator)</span>
  274. <span class="sd"> &gt;&gt;&gt; # operators list with crossover and mutation</span>
  275. <span class="sd"> &gt;&gt;&gt; operators = [SimpleCrossover(), SimpleMutation()]</span>
  276. <span class="sd"> &gt;&gt;&gt; policy = RandomPolicy(operators)</span>
  277. <span class="sd"> &gt;&gt;&gt; algo = IteratedLocalSearch(initializer, evaluator, operators, policy, validator, maximise=True, verbose=False)</span>
  278. <span class="sd"> &gt;&gt;&gt; # run the algorithm</span>
  279. <span class="sd"> &gt;&gt;&gt; solution = algo.run(100, ls_evaluations=10)</span>
  280. <span class="sd"> &gt;&gt;&gt; solution._score</span>
  281. <span class="sd"> 137</span>
  282. <span class="sd"> &quot;&quot;&quot;</span>
  283. <div class="viewcode-block" id="IteratedLocalSearch.run"><a class="viewcode-back" href="../../../macop/macop.algorithms.mono.html#macop.algorithms.mono.IteratedLocalSearch.run">[docs]</a> <span class="k">def</span> <span class="nf">run</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">evaluations</span><span class="p">,</span> <span class="n">ls_evaluations</span><span class="o">=</span><span class="mi">100</span><span class="p">):</span>
  284. <span class="sd">&quot;&quot;&quot;</span>
  285. <span class="sd"> Run the iterated local search algorithm using local search (EvE compromise)</span>
  286. <span class="sd"> Args:</span>
  287. <span class="sd"> evaluations: {int} -- number of global evaluations for ILS</span>
  288. <span class="sd"> ls_evaluations: {int} -- number of Local search evaluations (default: 100)</span>
  289. <span class="sd"> Returns:</span>
  290. <span class="sd"> {Solution} -- best solution found</span>
  291. <span class="sd"> &quot;&quot;&quot;</span>
  292. <span class="c1"># by default use of mother method to initialize variables</span>
  293. <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="n">run</span><span class="p">(</span><span class="n">evaluations</span><span class="p">)</span>
  294. <span class="c1"># enable resuming for ILS</span>
  295. <span class="bp">self</span><span class="o">.</span><span class="n">resume</span><span class="p">()</span>
  296. <span class="c1"># initialize current solution</span>
  297. <span class="bp">self</span><span class="o">.</span><span class="n">initRun</span><span class="p">()</span>
  298. <span class="c1"># passing global evaluation param from ILS</span>
  299. <span class="n">ls</span> <span class="o">=</span> <span class="n">HillClimberFirstImprovment</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">_initializer</span><span class="p">,</span>
  300. <span class="bp">self</span><span class="o">.</span><span class="n">_evaluator</span><span class="p">,</span>
  301. <span class="bp">self</span><span class="o">.</span><span class="n">_operators</span><span class="p">,</span>
  302. <span class="bp">self</span><span class="o">.</span><span class="n">_policy</span><span class="p">,</span>
  303. <span class="bp">self</span><span class="o">.</span><span class="n">_validator</span><span class="p">,</span>
  304. <span class="bp">self</span><span class="o">.</span><span class="n">_maximise</span><span class="p">,</span>
  305. <span class="n">verbose</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">_verbose</span><span class="p">,</span>
  306. <span class="n">parent</span><span class="o">=</span><span class="bp">self</span><span class="p">)</span>
  307. <span class="c1"># add same callbacks</span>
  308. <span class="k">for</span> <span class="n">callback</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">_callbacks</span><span class="p">:</span>
  309. <span class="n">ls</span><span class="o">.</span><span class="n">addCallback</span><span class="p">(</span><span class="n">callback</span><span class="p">)</span>
  310. <span class="c1"># local search algorithm implementation</span>
  311. <span class="k">while</span> <span class="ow">not</span> <span class="bp">self</span><span class="o">.</span><span class="n">stop</span><span class="p">():</span>
  312. <span class="c1"># create and search solution from local search</span>
  313. <span class="n">newSolution</span> <span class="o">=</span> <span class="n">ls</span><span class="o">.</span><span class="n">run</span><span class="p">(</span><span class="n">ls_evaluations</span><span class="p">)</span>
  314. <span class="c1"># if better solution than currently, replace it</span>
  315. <span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">isBetter</span><span class="p">(</span><span class="n">newSolution</span><span class="p">):</span>
  316. <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span> <span class="o">=</span> <span class="n">newSolution</span>
  317. <span class="c1"># number of evaluatins increased from LocalSearch</span>
  318. <span class="c1"># increase number of evaluations and progress are then not necessary there</span>
  319. <span class="c1">#self.increaseEvaluation()</span>
  320. <span class="c1">#self.progress()</span>
  321. <span class="bp">self</span><span class="o">.</span><span class="n">information</span><span class="p">()</span>
  322. <span class="n">logging</span><span class="o">.</span><span class="n">info</span><span class="p">(</span><span class="sa">f</span><span class="s2">&quot;End of </span><span class="si">{</span><span class="nb">type</span><span class="p">(</span><span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="vm">__name__</span><span class="si">}</span><span class="s2">, best solution found </span><span class="si">{</span><span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span><span class="si">}</span><span class="s2">&quot;</span><span class="p">)</span>
  323. <span class="bp">self</span><span class="o">.</span><span class="n">end</span><span class="p">()</span>
  324. <span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">_bestSolution</span></div></div>
  325. </pre></div>
  326. </div>
  327. </div>
  328. <footer>
  329. <hr/>
  330. <div role="contentinfo">
  331. <p>
  332. &copy; Copyright 2020, Jérôme BUISINE
  333. </p>
  334. </div>
  335. Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/rtfd/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>.
  336. </footer>
  337. </div>
  338. </div>
  339. </section>
  340. </div>
  341. <script type="text/javascript">
  342. jQuery(function () {
  343. SphinxRtdTheme.Navigation.enable(true);
  344. });
  345. </script>
  346. </body>
  347. </html>