Merge pull request #246 from ljvmiranda921/refactor/general

.coveragerc

@@ -0,0 +1,5 @@
+[report]
+exclude_lines =
+    pragma: no cover
+    @abc.abstractmethod
+    pos = init_pos

.isort.cfg

@@ -0,0 +1,5 @@
+[settings]
+skip=test_plotters.py
+import_heading_stdlib=Import standard library
+import_heading_firstparty=Import from pyswarms
+import_heading_thirdparty=Import modules

docs/conf.py

@@ -13,8 +13,12 @@
 # All configuration values have a default; values that are commented out
 # serve to show the default.
 
-import sys
+# Import standard library
 import os
+import sys
+
+# Import from pyswarms
+import pyswarms
 
 # If extensions (or modules to document with autodoc) are in another
 # directory, add these directories to sys.path here. If the directory is
@@ -32,7 +36,6 @@
 # sys.path.insert(0, project_root)
 sys.path.insert(0, os.path.abspath("../"))
 
-import pyswarms
 
 # -- General configuration ---------------------------------------------
 

pyswarms/backend/generators.py

@@ -9,8 +9,10 @@
 
 """
 
+# Import standard library
 import logging
 
+# Import modules
 import numpy as np
 
 from ..utils.reporter import Reporter
@@ -44,6 +46,14 @@ def generate_swarm(
     -------
     numpy.ndarray
         swarm matrix of shape (n_particles, n_dimensions)
+
+    Raises
+    ------
+    ValueError
+        When the shapes and values of bounds, dimensions, and init_pos
+        are inconsistent.
+    TypeError
+        When the argument passed to bounds is not an iterable.
     """
     try:
         if (init_pos is not None) and (bounds is None):
@@ -70,12 +80,13 @@ def generate_swarm(
                 low=min_bounds, high=max_bounds, size=(n_particles, dimensions)
             )
     except ValueError:
-        rep.logger.exception(
-            "Please check the size and value of bounds and dimensions"
-        )
+        msg = "Bounds and/or init_pos should be of size ({},)"
+        rep.logger.exception(msg.format(dimensions))
         raise
     except TypeError:
-        rep.logger.exception("Invalid input type!")
+        msg = "generate_swarm() takes an int for n_particles and dimensions and an array for bounds"
+        rep.logger.exception(msg)
+        raise
     else:
         return pos
 
@@ -96,6 +107,18 @@ def generate_discrete_swarm(
     init_pos : :code:`numpy.ndarray` (default is :code:`None`)
         option to explicitly set the particles' initial positions. Set to
         :code:`None` if you wish to generate the particles randomly.
+
+    Returns
+    -------
+    numpy.ndarray
+        swarm matrix of shape (n_particles, n_dimensions)
+
+    Raises
+    ------
+    ValueError
+        When init_pos during binary=True does not contain two unique values.
+    TypeError
+        When the argument passed to n_particles or dimensions is incorrect.
     """
     try:
         if (init_pos is not None) and binary:
@@ -112,11 +135,12 @@ def generate_discrete_swarm(
                 size=(n_particles, dimensions)
             ).argsort(axis=1)
     except ValueError:
-        rep.logger.exception(
-            "Please check the size and value of bounds and dimensions"
-        )
+        rep.logger.exception("Please check the size and value of dimensions")
+        raise
     except TypeError:
-        rep.logger.exception("Invalid input type!")
+        msg = "generate_discrete_swarm() takes an int for n_particles and dimensions"
+        rep.logger.exception(msg)
+        raise
     else:
         return pos
 
@@ -146,11 +170,13 @@ def generate_velocity(n_particles, dimensions, clamp=None):
             size=(n_particles, dimensions)
         ) + min_velocity
     except ValueError:
-        rep.logger.exception(
-            "Please check the size and value of clamp and dimensions"
-        )
+        msg = "Please check clamp shape: {} != {}"
+        rep.logger.exception(msg.format(len(clamp), dimensions))
+        raise
     except TypeError:
-        rep.logger.exception("Invalid input type!")
+        msg = "generate_velocity() takes an int for n_particles and dimensions and an array for clamp"
+        rep.logger.exception(msg)
+        raise
     else:
         return velocity
 

pyswarms/backend/operators.py

@@ -8,8 +8,10 @@
 to specify how the swarm will behave.
 """
 
+# Import standard library
 import logging
 
+# Import modules
 import numpy as np
 
 from ..utils.reporter import Reporter
@@ -69,6 +71,7 @@ def compute_pbest(swarm):
         rep.logger.exception(
             "Please pass a Swarm class. You passed {}".format(type(swarm))
         )
+        raise
     else:
         return (new_pbest_pos, new_pbest_cost)
 
@@ -139,8 +142,10 @@ def compute_velocity(swarm, clamp):
         rep.logger.exception(
             "Please pass a Swarm class. You passed {}".format(type(swarm))
         )
+        raise
     except KeyError:
         rep.logger.exception("Missing keyword in swarm.options")
+        raise
     else:
         return updated_velocity
 
@@ -187,5 +192,6 @@ def compute_position(swarm, bounds):
         rep.logger.exception(
             "Please pass a Swarm class. You passed {}".format(type(swarm))
         )
+        raise
     else:
         return position

pyswarms/backend/swarms.py

@@ -8,6 +8,7 @@
 as input to most backend cases.
 """
 
+# Import modules
 import numpy as np
 from attr import attrib, attrs
 from attr.validators import instance_of

pyswarms/backend/topology/base.py

@@ -12,6 +12,7 @@
 :mod:`pyswarms.backend.swarms.Swarm` module.
 """
 
+# Import standard library
 import abc
 import logging
 
@@ -33,7 +34,7 @@ def __init__(self, static, **kwargs):
             self.rep.log(
                 "Running on `dynamic` topology,"
                 "set `static=True` for fixed neighbors.",
-                lvl=10,
+                lvl=logging.DEBUG,
             )
 
     @abc.abstractmethod

pyswarms/backend/topology/pyramid.py

@@ -6,8 +6,10 @@
 This class implements a pyramid topology. In this topology, the particles are connected by N-dimensional simplices.
 """
 
+# Import standard library
 import logging
 
+# Import modules
 import numpy as np
 from scipy.spatial import Delaunay
 
@@ -29,7 +31,7 @@ def __init__(self, static=False):
         super(Pyramid, self).__init__(static)
         self.rep = Reporter(logger=logging.getLogger(__name__))
 
-    def compute_gbest(self, swarm):
+    def compute_gbest(self, swarm, **kwargs):
         """Update the global best using a pyramid neighborhood approach
 
         This topology uses the :code:`Delaunay` class from :code:`scipy`. To prevent precision errors in the Delaunay

pyswarms/backend/topology/random.py

@@ -6,9 +6,11 @@
 This class implements a random topology. All particles are connected in a random fashion.
 """
 
+# Import standard library
 import itertools
 import logging
 
+# Import modules
 import numpy as np
 from scipy.sparse.csgraph import connected_components, dijkstra
 
@@ -25,11 +27,12 @@ def __init__(self, static=False):
         ----------
         static : bool (Default is :code:`False`)
             a boolean that decides whether the topology
-            is static or dynamic"""
+            is static or dynamic
+        """
         super(Random, self).__init__(static)
         self.rep = Reporter(logger=logging.getLogger(__name__))
 
-    def compute_gbest(self, swarm, k):
+    def compute_gbest(self, swarm, k, **kwargs):
         """Update the global best using a random neighborhood approach
 
         This uses random class from :code:`numpy` to give every particle k
@@ -81,10 +84,14 @@ def compute_gbest(self, swarm, k):
             ).astype(int)
 
             # Obtain best cost and position
-            best_cost = np.min(swarm.pbest_cost[best_neighbor])
-            best_pos = swarm.pbest_pos[
-                best_neighbor[np.argmin(swarm.pbest_cost[best_neighbor])]
-            ]
+            if np.min(swarm.pbest_cost) < swarm.best_cost:
+                best_cost = np.min(swarm.pbest_cost[best_neighbor])
+                best_pos = swarm.pbest_pos[
+                    best_neighbor[np.argmin(swarm.pbest_cost[best_neighbor])]
+                ]
+            else:
+                # Just get the previous best_pos and best_cost
+                best_pos, best_cost = swarm.best_pos, swarm.best_cost
 
         except AttributeError:
             self.rep.logger.exception(

pyswarms/backend/topology/ring.py

@@ -9,8 +9,10 @@
 optimizers.
 """
 
+# Import standard library
 import logging
 
+# Import modules
 import numpy as np
 from scipy.spatial import cKDTree
 
@@ -27,11 +29,12 @@ def __init__(self, static=False):
         ----------
         static : bool (Default is :code:`False`)
             a boolean that decides whether the topology
-            is static or dynamic"""
+            is static or dynamic
+        """
         super(Ring, self).__init__(static)
         self.rep = Reporter(logger=logging.getLogger(__name__))
 
-    def compute_gbest(self, swarm, p, k):
+    def compute_gbest(self, swarm, p, k, **kwargs):
         """Update the global best using a ring-like neighborhood approach
 
         This uses the cKDTree method from :code:`scipy` to obtain the nearest
@@ -41,13 +44,13 @@ def compute_gbest(self, swarm, p, k):
         ----------
         swarm : pyswarms.backend.swarms.Swarm
             a Swarm instance
-        k : int
-            number of neighbors to be considered. Must be a
-            positive integer less than :code:`n_particles`
         p: int {1,2}
             the Minkowski p-norm to use. 1 is the
             sum-of-absolute values (or L1 distance) while 2 is
             the Euclidean (or L2) distance.
+        k : int
+            number of neighbors to be considered. Must be a
+            positive integer less than :code:`n_particles`
 
         Returns
         -------

pyswarms/backend/topology/star.py

@@ -9,8 +9,10 @@
 optimizers.
 """
 
+# Import standard library
 import logging
 
+# Import modules
 import numpy as np
 
 from .. import operators as ops
@@ -19,16 +21,17 @@
 
 
 class Star(Topology):
-    def __init__(self):
+    def __init__(self, static=None, **kwargs):
+        # static = None is just an artifact to make the API consistent
+        # Setting it will not change swarm behavior
         super(Star, self).__init__(static=True)
         self.rep = Reporter(logger=logging.getLogger(__name__))
 
-    def compute_gbest(self, swarm):
+    def compute_gbest(self, swarm, **kwargs):
         """Update the global best using a star topology
 
         This method takes the current pbest_pos and pbest_cost, then returns
-        the minimum cost and position from the matrix. It should be used in
-        tandem with an if statement
+        the minimum cost and position from the matrix.
 
         .. code-block:: python
 
@@ -39,10 +42,8 @@ def compute_gbest(self, swarm):
             my_swarm = P.create_swarm(n_particles, dimensions)
             my_topology = Star()
 
-            # If the minima of the pbest_cost is less than the best_cost
-            if np.min(pbest_cost) < best_cost:
-                # Update best_cost and position
-                swarm.best_pos, swarm.best_cost = my_topology.compute_best_particle(my_swarm)
+            # Update best_cost and position
+            swarm.best_pos, swarm.best_cost = my_topology.compute_gbest(my_swarm)
 
         Parameters
         ----------
@@ -61,8 +62,14 @@ def compute_gbest(self, swarm):
                 self.neighbor_idx = np.tile(
                     np.arange(swarm.n_particles), (swarm.n_particles, 1)
                 )
-            best_pos = swarm.pbest_pos[np.argmin(swarm.pbest_cost)]
-            best_cost = np.min(swarm.pbest_cost)
+            if np.min(swarm.pbest_cost) < swarm.best_cost:
+                # Get the particle position with the lowest pbest_cost
+                # and assign it to be the best_pos
+                best_pos = swarm.pbest_pos[np.argmin(swarm.pbest_cost)]
+                best_cost = np.min(swarm.pbest_cost)
+            else:
+                # Just get the previous best_pos and best_cost
+                best_pos, best_cost = swarm.best_pos, swarm.best_cost
         except AttributeError:
             self.rep.logger.exception(
                 "Please pass a Swarm class. You passed {}".format(type(swarm))