README.RST

README.rst

@@ -10,23 +10,32 @@
    :target: https://www.apache.org/licenses/LICENSE-2.0
 
 
+.. |logo| image:: doc/source/_theme/custom_theme/static/img/pymoo_banner_github.png
+  :target: https://pymoo.org
+  :alt: pymoo
+
+
+.. |animation| image:: doc/source/_theme/custom_theme/static/img/animation.gif
+  :target: https://pymoo.org
+  :alt: pymoo
+
+
 .. _Github: https://github.com/msu-coinlab/pymoo
 .. _Documentation: https://www.pymoo.org/
 .. _Paper: https://arxiv.org/abs/2002.04504
 
 
-.. raw:: html
 
-   <div>
-   <a href="https://pymoo.org"><img src="https://pymoo.org/_static/img/pymoo_banner_github.png" alt="pymoo"></a>
-   </div>
 
+|travis| |python| |license|
 
 
-|travis| |python| |license|
+|logo|
 
 
-Documentation_ / Paper_ / Installation_ / Usage_ / `Cite Us`_ / Contact_
+
+Documentation_ / Paper_ / Installation_ / Usage_ / Citation_ / Contact_
+
 
 
 
@@ -76,14 +85,14 @@ However, for instance executing NSGA2:
     from pymoo.optimize import minimize
     from pymoo.visualization.scatter import Scatter
 
-    problem = get_problem("zdt3")
+    problem = get_problem("zdt1")
 
     algorithm = NSGA2(pop_size=100)
 
     res = minimize(problem,
                    algorithm,
                    ('n_gen', 200),
-                   seed=84,
+                   seed=1,
                    verbose=True)
 
     plot = Scatter()
@@ -94,8 +103,12 @@ However, for instance executing NSGA2:
 
 
 
-.. _Cite Us:
-Cite Us
+|animation|
+
+
+
+.. _Citation:
+Citation
 ====================================================================
 
 We are currently working on a journal publication for *pymoo*.
@@ -125,3 +138,4 @@ Feel free to contact me if you have any question:
 | Computational Optimization and Innovation Laboratory (COIN)
 | East Lansing, MI 48824, USA
 
+

doc/github/README.template

@@ -10,23 +10,32 @@
    :target: https://www.apache.org/licenses/LICENSE-2.0
 
 
+.. |logo| image:: doc/source/_theme/custom_theme/static/img/pymoo_banner_github.png
+  :target: https://pymoo.org
+  :alt: pymoo
+
+
+.. |animation| image:: doc/source/_theme/custom_theme/static/img/animation.gif
+  :target: https://pymoo.org
+  :alt: pymoo
+
+
 .. _Github: https://github.com/msu-coinlab/pymoo
 .. _Documentation: https://www.pymoo.org/
 .. _Paper: https://arxiv.org/abs/2002.04504
 
 
-.. raw:: html
 
-   <div>
-   <a href="https://pymoo.org"><img src="https://pymoo.org/_static/img/pymoo_banner_github.png" alt="pymoo"></a>
-   </div>
 
+|travis| |python| |license|
 
 
-|travis| |python| |license|
+|logo|
 
 
-Documentation_ / Paper_ / Installation_ / Usage_ / `Cite Us`_ / Contact_
+
+Documentation_ / Paper_ / Installation_ / Usage_ / Citation_ / Contact_
+
 
 
 
@@ -76,8 +85,12 @@ However, for instance executing NSGA2:
 
 
 
-.. _Cite Us:
-Cite Us
+|animation|
+
+
+
+.. _Citation:
+Citation
 ====================================================================
 
 We are currently working on a journal publication for *pymoo*.
@@ -108,3 +121,4 @@ Feel free to contact me if you have any question:
 | East Lansing, MI 48824, USA
 
 
+

pymoo/model/plot.py

@@ -12,6 +12,8 @@
 class Plot:
 
     def __init__(self,
+                 fig=None,
+                 ax=None,
                  figsize=(8, 6),
                  title=None,
                  legend=False,
@@ -31,8 +33,8 @@ def __init__(self,
         plt.rc('font', family='serif')
 
         # the matplotlib classes
-        self.fig = None
-        self.ax = None
+        self.fig = fig
+        self.ax = ax
         self.figsize = figsize
 
         # the title of the figure - can also be a list if subfigures
@@ -81,6 +83,8 @@ def __init__(self,
         self.bounds = bounds
 
     def init_figure(self, n_rows=1, n_cols=1, plot_3D=False, force_axes_as_matrix=False):
+        if self.fig is not None or self.ax is not None:
+            return
 
         if not plot_3D:
             self.fig, self.ax = plt.subplots(nrows=n_rows, ncols=n_cols, figsize=self.figsize)

pymoo/model/population.py

@@ -12,6 +12,10 @@ def __new__(cls, n_individuals=0, individual=Individual()):
         obj.individual = individual
         return obj
 
+    @classmethod
+    def merge(cls, a, b):
+        return a.merge(b)
+
     def merge(self, other):
         if len(self) == 0:
             return other

pymoo/problems/single/multimodal.py

@@ -0,0 +1,26 @@
+import autograd.numpy as anp
+
+from pymoo.model.problem import Problem
+
+
+def curve(problem, n_points=200):
+    X = anp.linspace(problem.xl[0], problem.xu[0], n_points)[:, None]
+    F = problem.evaluate(X)
+    return anp.column_stack([X, F])
+
+
+class MultiModalSimple1(Problem):
+    def __init__(self, n_var=1):
+        super().__init__(n_var=n_var, n_obj=1, n_constr=0, xl=0, xu=1, type_var=anp.double)
+
+    def _evaluate(self, x, out, *args, **kwargs):
+        out["F"] = 1 - anp.exp(-x ** 2) * anp.sin(2 * anp.pi * x) ** 2
+
+
+class MultiModalSimple2(Problem):
+    def __init__(self, n_var=1):
+        super().__init__(n_var=n_var, n_obj=1, n_constr=0, xl=-1, xu=0, type_var=anp.double)
+
+    def _evaluate(self, x, out, *args, **kwargs):
+        x = - x
+        out["F"] = 1.1 - anp.exp(-2 * x) * anp.sin(5 * anp.pi * x) ** 2

pymoo/usage/algorithms/usage_ga.py

@@ -13,6 +13,3 @@
                termination=('n_gen', 50),
                seed=1,
                verbose=False)
-
-print("Best solution found: \nX = %s\nF = %s" % (res.X, res.F))
-

pymoo/usage/algorithms/usage_nsga2.py

@@ -3,14 +3,14 @@
 from pymoo.optimize import minimize
 from pymoo.visualization.scatter import Scatter
 
-problem = get_problem("zdt3")
+problem = get_problem("zdt1")
 
 algorithm = NSGA2(pop_size=100)
 
 res = minimize(problem,
                algorithm,
                ('n_gen', 200),
-               seed=84,
+               seed=1,
                verbose=True)
 
 plot = Scatter()

pymoo/usage/visualization/usage_gif.py

@@ -0,0 +1,48 @@
+import matplotlib.pyplot as plt
+
+from pymoo.algorithms.nsga2 import NSGA2
+from pymoo.optimize import minimize
+from pymoo.problems.multi import ZDT1
+from pymoo.util.nds.non_dominated_sorting import NonDominatedSorting
+from pymoo.visualization.pcp import PCP
+from pymoo.visualization.scatter import Scatter
+from pyrecorder.recorders.gif import GIF
+from pyrecorder.video import Video
+
+problem = ZDT1(n_var=6)
+
+algorithm = NSGA2()
+
+ret = minimize(problem,
+               algorithm,
+               termination=('n_gen', 61),
+               seed=1,
+               save_history=True,
+               verbose=False)
+
+print(ret.F)
+
+with Video(GIF("animation.gif")) as vid:
+    for algorithm in ret.history:
+
+        if algorithm.n_gen % 5 == 0:
+            X, F = algorithm.pop.get("X", "F")
+            nds = NonDominatedSorting().do(F, only_non_dominated_front=True)
+            other = [k for k in range(len(F)) if k not in nds]
+
+            fig, (ax1, ax2) = plt.subplots(2, figsize=(8, 6))
+            fig.suptitle("%s - %s - Gen %s" % ("ZDT1", "NSGA2", algorithm.n_gen), fontsize=16)
+
+            pcp = PCP(ax=ax1, bounds=(problem.xl, problem.xu))
+            pcp.set_axis_style(color="black", alpha=0.7)
+            pcp.add(X[other], color="blue", linewidth=0.5)
+            pcp.add(X[nds], color="red", linewidth=2)
+            pcp.do()
+
+            sc = Scatter(ax=ax2)
+            sc.add(F[other], color="blue")
+            sc.add(F[nds], color="red")
+            sc.add(problem.pareto_front(), plot_type="line", color="black")
+            sc.do()
+
+            vid.record()

pymoo/util/clearing.py

@@ -0,0 +1,58 @@
+import numpy as np
+
+
+def func_select_by_objective(pop):
+    F = pop.get("F")
+    return F[:, 0].argmin()
+
+
+def select_by_clearing(pop, D, n_select, func_select, eps=0.05):
+    clearing = EpsilonClearing(D, eps)
+
+    while len(clearing.selected()) < n_select:
+        remaining = clearing.remaining()
+
+        if len(remaining) == 0:
+            clearing.reset()
+
+        best = remaining[func_select(pop[remaining])]
+        clearing.select(best)
+
+    S = clearing.selected()
+    return S
+
+
+class EpsilonClearing:
+
+    def __init__(self, D,
+                 epsilon) -> None:
+        super().__init__()
+        self.D = D
+        self.n = len(D)
+        self.epsilon = epsilon
+
+        self.S = []
+        self.C = np.full(self.n, False)
+
+    def remaining(self):
+        return np.where(~self.C)[0]
+
+    def has_remaining(self):
+        return self.C.sum() != self.n
+
+    def cleared(self):
+        return self.C
+
+    def selected(self):
+        return self.S
+
+    def reset(self):
+        self.C = np.full(self.n, False)
+        self.C[self.S] = True
+
+    def select(self, k):
+        self.S.append(k)
+        self.C[k] = True
+
+        less_than_epsilon = self.D[k] < self.epsilon
+        self.C[less_than_epsilon] = True

pymoo/util/misc.py

@@ -90,15 +90,23 @@ def cdist(A, B, **kwargs):
     return scipy.spatial.distance.cdist(A, B, **kwargs)
 
 
-def vectorized_cdist(A, B, func_dist=euclidean_distance, **kwargs):
+def vectorized_cdist(A, B, func_dist=euclidean_distance, fill_diag_with_inf=False, **kwargs):
     u = np.repeat(A, B.shape[0], axis=0)
     v = np.tile(B, (A.shape[0], 1))
 
     D = func_dist(u, v, **kwargs)
     M = np.reshape(D, (A.shape[0], B.shape[0]))
+
+    if fill_diag_with_inf:
+        np.fill_diagonal(M, np.inf)
+
     return M
 
 
+def norm_eucl_dist(problem, A, B, **kwargs):
+    return vectorized_cdist(A, B, func_dist=norm_euclidean_distance(problem), **kwargs)
+
+
 def covert_to_type(problem, X):
     if problem.type_var == np.double:
         return X.astype(np.double)

pymoo/util/roulette.py

@@ -0,0 +1,27 @@
+import numpy as np
+
+
+class RouletteWheelSelection:
+
+    def __init__(self, val, larger_is_better=True):
+        super().__init__()
+        if not larger_is_better:
+            val = val.max() - val
+        _sum = val.sum()
+        self.cumulative = np.array([val[:k].sum() / _sum for k in range(1, len(val))])
+
+    def next(self, n=None):
+        if n is None:
+            X = np.random.random((1, 1))
+        else:
+            X = np.random.random((n, 1))
+            if n > 1:
+                X.repeat(n - 1, axis=1)
+
+        M = self.cumulative[None, :].repeat(len(X), axis=0)
+        B = X >= M
+        ret = B.sum(axis=1)
+
+        if n is None:
+            return ret[0]
+        return ret