Add docs for v2.0.0a4 on GitHub Pages

docs/pre/.buildinfo

@@ -1,4 +1,4 @@
 # Sphinx build info version 1
 # This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done.
-config: cded3a427bff386c9945aacc676c5163
+config: b8b49b2c9835f2e65ccc23820533afe5
 tags: 645f666f9bcd5a90fca523b33c5a78b7

docs/pre/README.html

@@ -5,11 +5,11 @@
     <meta charset="utf-8" />
     <meta name="viewport" content="width=device-width, initial-scale=1.0" /><meta name="viewport" content="width=device-width, initial-scale=1" />
 
-    <title>Contributors &#8212; lincs 2.0.0a3 documentation</title>
+    <title>Contributors &#8212; lincs 2.0.0a4 documentation</title>
     <link rel="stylesheet" type="text/css" href="_static/pygments.css?v=4f649999" />
     <link rel="stylesheet" type="text/css" href="_static/alabaster.css?v=039e1c02" />
     <link rel="stylesheet" type="text/css" href="_static/graphviz.css?v=eafc0fe6" />
-    <script data-url_root="./" id="documentation_options" src="_static/documentation_options.js?v=50df3787"></script>
+    <script data-url_root="./" id="documentation_options" src="_static/documentation_options.js?v=9b56a74a"></script>
     <script src="_static/doctools.js?v=888ff710"></script>
     <script src="_static/sphinx_highlight.js?v=4825356b"></script>
     <link rel="index" title="Index" href="genindex.html" />

docs/pre/_downloads/019271a03d05a599bcfb6d68c868c86a/python-api.ipynb

@@ -378,7 +378,7 @@
     "profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)\n",
     "weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)\n",
     "profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)\n",
-    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)\n",
+    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)\n",
     "termination_strategy = lc.TerminateAtAccuracy(models_being_learned, target_accuracy=len(learning_set.alternatives))"
    ]
   },
@@ -1970,7 +1970,7 @@
     "profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)\n",
     "weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)\n",
     "profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)\n",
-    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)\n",
+    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)\n",
     "termination_strategy = lc.TerminateAfterIterations(models_being_learned, max_iterations_count=1)\n",
     "\n",
     "lc.LearnMrsortByWeightsProfilesBreed(\n",
@@ -2032,7 +2032,7 @@
     {
      "data": {
       "text/plain": [
-       "[728, 593, 725, 671, 637, 609, 676, 833, 640]"
+       "[764, 595, 725, 666, 637, 609, 676, 830, 640]"
       ]
      },
      "execution_count": 72,
@@ -2127,14 +2127,14 @@
     {
      "data": {
       "text/plain": [
-       "[[1.0132789611816406e-06, 0.9999989867210388, 1.0132789611816406e-06],\n",
-       " [0.9999989867210388, 0.9999989867210388, 1.0132789611816406e-06],\n",
+       "[[0.5, 0.5, 0.5],\n",
+       " [1.0, 0.9999989867210388, 1.0132789611816406e-06],\n",
+       " [1.0, 0.0, 0.9999989867210388],\n",
        " [1.0, 0.0, 0.9999989867210388],\n",
-       " [0.5000004768371582, 0.4999994933605194, 0.4999994933605194],\n",
        " [1.0132789611816406e-06, 0.9999989867210388, 0.0],\n",
-       " [0.9999989867210388, 0.0, 1.0132789611816406e-06],\n",
-       " [1.0, 0.0, 0.0],\n",
-       " [1.0, 0.9999989867210388, 1.0132789611816406e-06],\n",
+       " [1.0132789611816406e-06, 0.0, 0.9999989867210388],\n",
+       " [1.0, 0.0, 0.9999989867210388],\n",
+       " [1.0, 0.0, 1.0],\n",
        " [0.9999989867210388, 0.0, 0.9999989867210388]]"
       ]
      },
@@ -2177,7 +2177,7 @@
       "sufficient_coalitions:\n",
       "  - &coalitions\n",
       "    kind: weights\n",
-      "    criterion_weights: [1, 0.999998987, 1.01327896e-06]\n",
+      "    criterion_weights: [1, 0, 1]\n",
       "  - *coalitions\n"
      ]
     }
@@ -2195,7 +2195,7 @@
     {
      "data": {
       "text/plain": [
-       "833"
+       "830"
       ]
      },
      "execution_count": 77,
@@ -2281,7 +2281,7 @@
     "profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)\n",
     "weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)\n",
     "profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)\n",
-    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)\n",
+    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)\n",
     "termination_strategy = lc.TerminateAtAccuracy(models_being_learned, target_accuracy=len(learning_set.alternatives))\n",
     "observer = VerboseObserver(models_being_learned)\n",
     "\n",
@@ -2327,7 +2327,7 @@
       "sufficient_coalitions:\n",
       "  - &coalitions\n",
       "    kind: weights\n",
-      "    criterion_weights: [1, 0.999998987, 1.01327896e-06]\n",
+      "    criterion_weights: [1, 0, 1]\n",
       "  - *coalitions\n",
       "kind: ncs-classification-model\n",
       "format_version: 1\n",
@@ -2361,23 +2361,23 @@
       "format_version: 1\n",
       "accepted_values:\n",
       "  - kind: thresholds\n",
-      "    thresholds: [7.96338844, 3.74707603]\n",
+      "    thresholds: [7.93314505, 3.74707603]\n",
       "  - kind: thresholds\n",
-      "    thresholds: [94, 99]\n",
+      "    thresholds: [9, 9]\n",
       "  - kind: thresholds\n",
       "    thresholds: [B, B]\n",
       "sufficient_coalitions:\n",
       "  - &coalitions\n",
       "    kind: weights\n",
-      "    criterion_weights: [1, 0, 1]\n",
+      "    criterion_weights: [1, 0.999998987, 1]\n",
       "  - *coalitions\n",
       "kind: ncs-classification-model\n",
       "format_version: 1\n",
       "accepted_values:\n",
       "  - kind: thresholds\n",
       "    thresholds: [7.95116329, 3.74707603]\n",
       "  - kind: thresholds\n",
-      "    thresholds: [94, 99]\n",
+      "    thresholds: [73, 99]\n",
       "  - kind: thresholds\n",
       "    thresholds: [B, B]\n",
       "sufficient_coalitions:\n",
@@ -2410,7 +2410,7 @@
     "profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)\n",
     "weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)\n",
     "profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)\n",
-    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)\n",
+    "breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)\n",
     "termination_strategy = lc.TerminateAtAccuracy(models_being_learned, target_accuracy=len(learning_set.alternatives))\n",
     "observer = IntermediatesObserver(problem, models_being_learned)\n",
     "\n",
@@ -2634,18 +2634,19 @@
       "initialize_profiles 0 9\n",
       "optimize_weights 0 9\n",
       "improve_profiles 0 9\n",
+      "optimize_weights 0 9\n",
       "terminate\n",
       "breed\n",
-      "optimize_weights 0 9\n",
       "improve_profiles 0 9\n",
+      "optimize_weights 0 9\n",
       "terminate\n",
       "breed\n",
-      "optimize_weights 0 9\n",
       "improve_profiles 0 9\n",
+      "optimize_weights 0 9\n",
       "terminate\n",
       "breed\n",
-      "optimize_weights 0 9\n",
       "improve_profiles 0 9\n",
+      "optimize_weights 0 9\n",
       "terminate\n"
      ]
     }

docs/pre/_sources/changelog.rst.txt

@@ -4,6 +4,17 @@
 Changelog
 =========
 
+Version 2.0.0a4 (2024-11-08)
+============================
+
+- **Changes behavior slightly** Optimize WPB model weights before evaluating and returning them
+- **Breaking** Add parameter ``weights_optimization_strategy`` to the ``ReinitializeLeastAccurate`` breeding strategy (required to fix previous point)
+- **Changes behavior slightly** Always keep the best model during WPB
+- Fix the ``universal2`` wheels for macOS to actually run on M1 Macs
+- **Changes behavior slightly** Upgrade OR-Tools from 8.2 to 9.11 on Linux and macOS (required for previous point)
+
+**Note** On Windows, the package still uses OR-Tools 8.2.
+
 Version 2.0.0a3 (2024-10-10)
 ============================
 
@@ -14,8 +25,8 @@ Version 2.0.0a3 (2024-10-10)
     - exposed on the command-line with the ``--mrsort.weights-profiles-breed.linear-program.solver`` and ``--mrsort.weights-profiles-breed.linear-program.experimental-in-house-simplex.processor`` options
     - very EXPERIMENTAL, with unsatisfactory results for now
 
-Version 2.0.0 (2024-XX-YY)
-==========================
+Version 2.0.0 (planned 2024-12-05)
+==================================
 
 - Support Python 3.12
 

docs/pre/_sources/get-started.rst.txt

@@ -73,7 +73,7 @@ The generated ``problem.yml`` should look like:
 
 .. code:: yaml
 
-    # Reproduction command (with lincs version 2.0.0a3): lincs generate classification-problem 4 3 --random-seed 40
+    # Reproduction command (with lincs version 2.0.0a4): lincs generate classification-problem 4 3 --random-seed 40
     kind: classification-problem
     format_version: 1
     criteria:
@@ -144,7 +144,7 @@ It should look like:
 
 .. code:: yaml
 
-    # Reproduction command (with lincs version 2.0.0a3): lincs generate classification-model problem.yml --random-seed 41 --model-type mrsort
+    # Reproduction command (with lincs version 2.0.0a4): lincs generate classification-model problem.yml --random-seed 41 --model-type mrsort
     kind: ncs-classification-model
     format_version: 1
     accepted_values:
@@ -217,7 +217,7 @@ It should start with something like this, and contain 1000 alternatives:
 
 .. code:: text
 
-    # Reproduction command (with lincs version 2.0.0a3): lincs generate classified-alternatives problem.yml model.yml 1000 --random-seed 42 --misclassified-count 0
+    # Reproduction command (with lincs version 2.0.0a4): lincs generate classified-alternatives problem.yml model.yml 1000 --random-seed 42 --misclassified-count 0
     name,"Criterion 1","Criterion 2","Criterion 3","Criterion 4",category
     "Alternative 1",0.37454012,0.796543002,0.95071429,0.183434784,"Best category"
     "Alternative 2",0.731993914,0.779690981,0.598658502,0.596850157,"Intermediate category 1"
@@ -253,7 +253,7 @@ so it is numerically different:
 
 .. code:: yaml
 
-    # Reproduction command (with lincs version 2.0.0a3): lincs learn classification-model problem.yml learning-set.csv --model-type mrsort --mrsort.strategy weights-profiles-breed --mrsort.weights-profiles-breed.models-count 9 --mrsort.weights-profiles-breed.accuracy-heuristic.random-seed 43 --mrsort.weights-profiles-breed.initialization-strategy maximize-discrimination-per-criterion --mrsort.weights-profiles-breed.weights-strategy linear-program --mrsort.weights-profiles-breed.linear-program.solver glop --mrsort.weights-profiles-breed.profiles-strategy accuracy-heuristic --mrsort.weights-profiles-breed.accuracy-heuristic.processor cpu --mrsort.weights-profiles-breed.breed-strategy reinitialize-least-accurate --mrsort.weights-profiles-breed.reinitialize-least-accurate.portion 0.5 --mrsort.weights-profiles-breed.target-accuracy 1.0
+    # Reproduction command (with lincs version 2.0.0a4): lincs learn classification-model problem.yml learning-set.csv --model-type mrsort --mrsort.strategy weights-profiles-breed --mrsort.weights-profiles-breed.models-count 9 --mrsort.weights-profiles-breed.accuracy-heuristic.random-seed 43 --mrsort.weights-profiles-breed.initialization-strategy maximize-discrimination-per-criterion --mrsort.weights-profiles-breed.weights-strategy linear-program --mrsort.weights-profiles-breed.linear-program.solver glop --mrsort.weights-profiles-breed.profiles-strategy accuracy-heuristic --mrsort.weights-profiles-breed.accuracy-heuristic.processor cpu --mrsort.weights-profiles-breed.breed-strategy reinitialize-least-accurate --mrsort.weights-profiles-breed.reinitialize-least-accurate.portion 0.5 --mrsort.weights-profiles-breed.target-accuracy 1.0
     kind: ncs-classification-model
     format_version: 1
     accepted_values:

docs/pre/_sources/python-api.md.txt

@@ -208,7 +208,7 @@ models_being_learned = lc.LearnMrsortByWeightsProfilesBreed.ModelsBeingLearned(p
 profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)
 weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)
 profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)
-breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)
+breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)
 termination_strategy = lc.TerminateAtAccuracy(models_being_learned, target_accuracy=len(learning_set.alternatives))
 ```
 
@@ -1146,7 +1146,7 @@ models_being_learned = lc.LearnMrsortByWeightsProfilesBreed.ModelsBeingLearned(p
 profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)
 weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)
 profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)
-breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)
+breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)
 termination_strategy = lc.TerminateAfterIterations(models_being_learned, max_iterations_count=1)
 
 lc.LearnMrsortByWeightsProfilesBreed(
@@ -1189,7 +1189,7 @@ list(models_being_learned.accuracies)  # Indexed by model_index
 
 
 ```text
-[728, 593, 725, 671, 637, 609, 676, 833, 640]
+[764, 595, 725, 666, 637, 609, 676, 830, 640]
 ```
 
 
@@ -1245,14 +1245,14 @@ Its `weights` attribute holds the MR-Sort weight of each criterion in each in-pr
 
 
 ```text
-[[1.0132789611816406e-06, 0.9999989867210388, 1.0132789611816406e-06],
- [0.9999989867210388, 0.9999989867210388, 1.0132789611816406e-06],
+[[0.5, 0.5, 0.5],
+ [1.0, 0.9999989867210388, 1.0132789611816406e-06],
+ [1.0, 0.0, 0.9999989867210388],
  [1.0, 0.0, 0.9999989867210388],
- [0.5000004768371582, 0.4999994933605194, 0.4999994933605194],
  [1.0132789611816406e-06, 0.9999989867210388, 0.0],
- [0.9999989867210388, 0.0, 1.0132789611816406e-06],
- [1.0, 0.0, 0.0],
- [1.0, 0.9999989867210388, 1.0132789611816406e-06],
+ [1.0132789611816406e-06, 0.0, 0.9999989867210388],
+ [1.0, 0.0, 0.9999989867210388],
+ [1.0, 0.0, 1.0],
  [0.9999989867210388, 0.0, 0.9999989867210388]]
 ```
 
@@ -1278,7 +1278,7 @@ accepted_values:
 sufficient_coalitions:
   - &coalitions
     kind: weights
-    criterion_weights: [1, 0.999998987, 1.01327896e-06]
+    criterion_weights: [1, 0, 1]
   - *coalitions
 ```
 
@@ -1292,7 +1292,7 @@ models_being_learned.get_best_accuracy()
 
 
 ```text
-833
+830
 ```
 
 
@@ -1327,7 +1327,7 @@ We can now pass it to a learning and perform that learning to observe its effect
 profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)
 weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)
 profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)
-breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)
+breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)
 termination_strategy = lc.TerminateAtAccuracy(models_being_learned, target_accuracy=len(learning_set.alternatives))
 observer = VerboseObserver(models_being_learned)
 
@@ -1381,7 +1381,7 @@ models_being_learned = lc.LearnMrsortByWeightsProfilesBreed.ModelsBeingLearned(p
 profiles_initialization_strategy = lc.InitializeProfilesForProbabilisticMaximalDiscriminationPowerPerCriterion(preprocessed_learning_set, models_being_learned)
 weights_optimization_strategy = lc.OptimizeWeightsUsingGlop(preprocessed_learning_set, models_being_learned)
 profiles_improvement_strategy = lc.ImproveProfilesWithAccuracyHeuristicOnCpu(preprocessed_learning_set, models_being_learned)
-breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy=profiles_initialization_strategy, count=4)
+breeding_strategy = lc.ReinitializeLeastAccurate(models_being_learned, profiles_initialization_strategy, weights_optimization_strategy, count=4)
 termination_strategy = lc.TerminateAtAccuracy(models_being_learned, target_accuracy=len(learning_set.alternatives))
 observer = IntermediatesObserver(problem, models_being_learned)
 
@@ -1414,7 +1414,7 @@ accepted_values:
 sufficient_coalitions:
   - &coalitions
     kind: weights
-    criterion_weights: [1, 0.999998987, 1.01327896e-06]
+    criterion_weights: [1, 0, 1]
   - *coalitions
 kind: ncs-classification-model
 format_version: 1
@@ -1448,23 +1448,23 @@ kind: ncs-classification-model
 format_version: 1
 accepted_values:
   - kind: thresholds
-    thresholds: [7.96338844, 3.74707603]
+    thresholds: [7.93314505, 3.74707603]
   - kind: thresholds
-    thresholds: [94, 99]
+    thresholds: [9, 9]
   - kind: thresholds
     thresholds: [B, B]
 sufficient_coalitions:
   - &coalitions
     kind: weights
-    criterion_weights: [1, 0, 1]
+    criterion_weights: [1, 0.999998987, 1]
   - *coalitions
 kind: ncs-classification-model
 format_version: 1
 accepted_values:
   - kind: thresholds
     thresholds: [7.95116329, 3.74707603]
   - kind: thresholds
-    thresholds: [94, 99]
+    thresholds: [73, 99]
   - kind: thresholds
     thresholds: [B, B]
 sufficient_coalitions:
@@ -1605,18 +1605,19 @@ learned_model = lc.LearnMrsortByWeightsProfilesBreed(
 initialize_profiles 0 9
 optimize_weights 0 9
 improve_profiles 0 9
+optimize_weights 0 9
 terminate
 breed
-optimize_weights 0 9
 improve_profiles 0 9
+optimize_weights 0 9
 terminate
 breed
-optimize_weights 0 9
 improve_profiles 0 9
+optimize_weights 0 9
 terminate
 breed
-optimize_weights 0 9
 improve_profiles 0 9
+optimize_weights 0 9
 terminate
 ```
 

docs/pre/_sources/reference/lincs.rst.txt

@@ -1040,7 +1040,7 @@
 
             The breeding strategy described in Olivier Sobrie's PhD thesis: re-initializes ``count`` in-progress models.
 
-            .. method:: __init__(models_being_learned: ModelsBeingLearned, profiles_initialization_strategy: ProfilesInitializationStrategy, count: int)
+            .. method:: __init__(models_being_learned: ModelsBeingLearned, profiles_initialization_strategy: ProfilesInitializationStrategy, weights_optimization_strategy: WeightsOptimizationStrategy, count: int)
 
                 Constructor. Keeps references to the profiles initialization strategy and the learning data.