- docstrings and tidying.

Author: andrewtliew

src/compas/numerical/__init__.py

@@ -38,8 +38,6 @@
     descent
     devo_numpy
     ga
-    lma
-    mma
     moga
 
 

src/compas/numerical/solvers/descent.py

@@ -1,5 +1,6 @@
-from __future__ import print_function
 from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
 
 import sys
 
@@ -25,10 +26,10 @@
     e = sqrt(eps)
 
 
-__author__     = ['Andrew Liew <[email protected]>']
-__copyright__  = 'Copyright 2017, BLOCK Research Group - ETH Zurich'
-__license__    = 'MIT License'
-__email__      = '[email protected]'
+__author__    = ['Andrew Liew <[email protected]>']
+__copyright__ = 'Copyright 2017, BLOCK Research Group - ETH Zurich'
+__license__   = 'MIT License'
+__email__     = '[email protected]'
 
 
 __all__ = [
@@ -39,7 +40,8 @@
 def descent(x0, fn, iterations=1000, gtol=10**(-6), bounds=None, limit=0, args=()):
     """A gradient descent optimisation solver.
 
-    Parameters:
+    Parameters
+    ----------
         x0 (array, list): n x 1 starting guess of x.
         fn (obj): The objective function to minimise.
         iterations (int): Maximum number of iterations.
@@ -48,21 +50,24 @@ def descent(x0, fn, iterations=1000, gtol=10**(-6), bounds=None, limit=0, args=(
         limit (float): Value of the objective function for which to terminate optimisation.
         args (tuple): Additional parameters needed for fn.
 
-    Returns:
+    Returns
+    -------
         float: Final value of the objective function.
         array: Values of x at the found local minimum.
     """
     r = 0.5
     c = 0.0001
     n = len(x0)
     x0 = reshape(array(x0), (n, 1))
+
     if bounds:
         bounds = array(bounds)
         lb = bounds[:, 0][:, newaxis]
         ub = bounds[:, 1][:, newaxis]
     else:
         lb = ones((n, 1)) * -10**20
         ub = ones((n, 1)) * +10**20
+
     zn = zeros((n, 1))
     g = zeros((n, 1))
     v = eye(n) * e
@@ -74,26 +79,32 @@ def phi(x, mu, *args):
     i = 0
     mu = 1
     while i < iterations:
+
         p0 = phi(x0, mu, *args)
         for j in range(n):
             vj = v[:, j][:, newaxis]
             g[j, 0] = (phi(x0 + vj, mu, *args) - p0) / e
         D = sum(-g * g)
+
         a = 1
         x1 = x0 - a * g
         while phi(x1, mu, *args) > p0 + c * a * D:
             a *= r
             x1 = x0 - a * g
         x0 -= a * g
+
         mu *= 10
         res = mean(abs(g))
         i += 1
         f1 = phi(x0, mu, *args)
-        print('Iteration: {0}  fopt: {1:.3g}  gres: {2:.3g}  step: {3}'.format(i, f1, res, a))
+
         if f1 < limit:
             break
         if res < gtol:
             break
+
+        print('Iteration: {0}  fopt: {1:.3g}  gres: {2:.3g}  step: {3}'.format(i, f1, res, a))
+
     return f1, x0
 
 

src/compas/numerical/solvers/devo_numpy.py

@@ -1,5 +1,6 @@
-from __future__ import print_function
 from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
 
 import sys
 
@@ -28,10 +29,10 @@
 import json
 
 
-__author__     = ['Andrew Liew <[email protected]>']
-__copyright__  = 'Copyright 2017, BLOCK Research Group - ETH Zurich'
-__license__    = 'MIT License'
-__email__      = '[email protected]'
+__author__    = ['Andrew Liew <[email protected]>']
+__copyright__ = 'Copyright 2017, BLOCK Research Group - ETH Zurich'
+__license__   = 'MIT License'
+__email__     = '[email protected]'
 
 
 __all__ = [
@@ -43,10 +44,8 @@ def devo_numpy(fn, bounds, population, generations, limit=0, results=None, vecto
                args=(), callback=None, **kwargs):
     """Call the Differential Evolution solver.
 
-    Note:
-        fn must return vectorised output for input (k, population) if vectored is True.
-
-    Parameters:
+    Parameters
+    ----------
         fn (obj): The function to evaluate and minimise.
         bounds (list): Lower and upper bounds for each DoF [[lb, ub], ...].
         population (int): Number of agents in the population.
@@ -60,11 +59,15 @@ def devo_numpy(fn, bounds, population, generations, limit=0, results=None, vecto
         args (seq): Sequence of optional arguments to pass to fn.
         callback (obj): Callback function for each generation.
 
-    Returns:
+    Returns
+    -------
         float: Optimal value of objective function.
         list: Values that give the optimum (minimised) function.
-    """
 
+    Note
+    ----
+        fn must return vectorised output for input (k, population) if vectored is True.
+    """
     tic = time()
 
     # Heading

src/compas/numerical/solvers/ga.py

@@ -1,16 +1,17 @@
-from __future__ import print_function
 from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
 
 import re
 import random
 import json
 import copy
 
 
-__author__     = ['Tomas Mendez Echenagucia <[email protected]>']
-__copyright__  = 'Copyright 2017, Block Research Group - ETH Zurich'
-__license__    = 'MIT License'
-__email__      = '[email protected]'
+__author__    = ['Tomas Mendez Echenagucia <[email protected]>']
+__copyright__ = 'Copyright 2017, Block Research Group - ETH Zurich'
+__license__   = 'MIT License'
+__email__     = '[email protected]'
 
 
 __all__ = [
@@ -60,7 +61,6 @@ def ga(fit_function,
        output_path=None,
        input_path=None):
 
-    ga_ = GA()
     """Genetic Algorithm optimisation (Holland 1975).
 
     Parameters
@@ -181,6 +181,7 @@ def foo(X):
         ================================================================================
 
     """
+    ga_ = GA()
 
     ga_.fit_name             = fit_name or fit_function.__name__
     ga_.fit_type             = fit_type
@@ -448,7 +449,6 @@ def scale_population(self, decoded_pop):
         scaled_pop: list
             The scaled ppopulation list.
         """
-
         scaled_pop = [[[]] * self.num_var for i in range(self.num_pop)]
         for j in range(self.num_pop):
             for i in range(self.num_var):
@@ -496,7 +496,6 @@ def select_elite_pop(self, pop, num_elite=None):
         elite_pop: dict
             The elite population dictionary.
         """
-
         if self.fit_type == 'min':
             sorted_ind = self.get_sorting_indices(self.current_pop['fit_value'])
         elif self.fit_type == 'max':
@@ -565,7 +564,6 @@ def simple_crossover(self):
         combined with individuals in ``GA.mating_pool_b`` using a single, randomly selected
         crossover point.
         """
-
         self.current_pop  = {'binary': [], 'decoded': [], 'scaled': [], 'fit_value': []}
         self.current_pop['binary'] = [[[]] * self.num_var for i in range(self.num_pop)]
         for j in range((self.num_pop - self.num_elite) / 2):
@@ -603,7 +601,7 @@ def code_decoded(self, decoded_pop):
         Parameters
         ----------
         decoded_pop: dict
-        The decoded population dictionary to be coded
+            The decoded population dictionary to be coded
 
         Returns
         -------
@@ -844,6 +842,10 @@ def get_best_individual_index(self):
         self.best_individual_index = indices[0]
 
 
+# ==============================================================================
+# Main
+# ==============================================================================
+
 if __name__ == '__main__':
 
     import os

src/compas/numerical/solvers/lma.py

@@ -1,7 +1,10 @@
-""""""
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
 
 __author__    = ['Tom Van Mele', ]
-__copyright__ = 'Copyright 2016 - Block Research Group, ETH Zurich'
+__copyright__ = 'Copyright 2017 - Block Research Group, ETH Zurich'
 __license__   = 'MIT License'
 __email__     = '[email protected]'
 

src/compas/numerical/solvers/mma.py

@@ -1,5 +1,10 @@
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+
 __author__    = ['Tom Van Mele', ]
-__copyright__ = 'Copyright 2016 - Block Research Group, ETH Zurich'
+__copyright__ = 'Copyright 2017 - Block Research Group, ETH Zurich'
 __license__   = 'MIT License'
 __email__     = '[email protected]'
 

src/compas/numerical/solvers/moga.py

@@ -1,15 +1,16 @@
-from __future__ import print_function
 from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
 
 import re
 import random
 import json
 
 
-__author__     = ['Tomas Mendez Echenagucia <[email protected]>']
-__copyright__  = 'Copyright 2016, Block Research Group - ETH Zurich'
-__license__    = 'MIT License'
-__email__      = '[email protected]'
+__author__    = ['Tomas Mendez Echenagucia <[email protected]>']
+__copyright__ = 'Copyright 2017, Block Research Group - ETH Zurich'
+__license__   = 'MIT License'
+__email__     = '[email protected]'
 
 
 __all__ = [
@@ -54,11 +55,10 @@ def moga(fit_functions,
          fkwargs=None,
          output_path=None):
 
-    """Genetic Algorithm optimisation [deb2001]_.
+    """Multi-objective Genetic Algorithm optimisation [deb2001]_.
 
     Parameters
     ----------
-
     fit_functions : list
         List of functions to be used by the :class'MOGA' to determine the fitness values.
         The function must have as a first argument a list of variables that determine the
@@ -471,7 +471,6 @@ def decode_binary_pop(self, bin_pop):
         return decoded_pop
 
     def scale_population(self, decoded_pop):
-
         """Scales the decoded population, variable values are scaled according to each
         of their bounds contained in ``GA.boundaries``.
 
@@ -485,7 +484,6 @@ def scale_population(self, decoded_pop):
         scaled_pop: list
             The scaled ppopulation list.
         """
-
         scaled_pop = [[[]] * self.num_var for i in range(self.num_pop)]
         for j in range(self.num_pop):
             for i in range(self.num_var):
@@ -802,7 +800,6 @@ def get_pop_from_pf_file(self):
         lines = pf_file.readlines()
         pf_file.close()
 
-
         file_pop['scaled'] = [[[]] * self.num_var for i in range(self.num_pop)]
         file_pop['scaled'] = [[[]] * self.num_fit_func for i in range(self.num_pop)]
         for i in range(self.num_pop):
@@ -923,7 +920,6 @@ def make_pop_pf_dict(self):
         -------
         pf_dict: dict
             The dictionary containing the pareto front level of all individuals.
-
         """
         pf_dict = {}
         for j in range(len(self.pareto_front_indices) - 1):

src/compas/utilities/animation.py

@@ -3,8 +3,10 @@
 from __future__ import division
 
 import os
-import imageio
-
+try:
+    import imageio
+except:
+    pass
 
 __author__    = ['Matthias Rippmann', ]
 __copyright__ = 'Copyright 2016 - Block Research Group, ETH Zurich'