Fixed ackley function

Author: HaaLeo

CHANGELOG.md

@@ -3,6 +3,9 @@ All notable changes to the "swarmlib" pypi package will be documented in this fi
 This project follows [semantic versioning](https://semver.org/).
 
 
+## 2020-01-09 - v0.3.2
+* **Fixed** a bug in the firefly algorithm that caused the application to crash when the ackley function was selected.
+
 ## 2019-10-30 - v0.3.1
 * **Fixed** a bug in the ACO algorithm that chose the next node by its maximal attractiveness. Now the next node is chosen randomly weighted by its attractiveness
 

README.md

@@ -34,7 +34,7 @@ The implementation was part of the course [Natural computing for learning and op
 
 ### Features
 
-Enables to apply the ant colony optimization algorithm to a TSP using a [TSPLIB95](https://www.iwr.uni-heidelberg.de/groups/comopt/software/TSPLIB95/index.html) file and plots the result.
+Enables to apply the ant colony optimization algorithm to a TSP using a [TSPLIB95](http://comopt.ifi.uni-heidelberg.de/software/TSPLIB95/) file and plots the result.
 
 ![ACO Sample](https://raw.githubusercontent.com/HaaLeo/swarmlib/master/doc/ACO_Sample.png)
 

swarmlib/_version.py

@@ -3,4 +3,4 @@
 #  Licensed under the BSD 3-Clause License. See LICENSE.txt in the project root for license information.
 # ------------------------------------------------------------------------------------------------------
 
-__version__ = '0.3.1'
+__version__ = '0.3.2'

swarmlib/fireflyalgorithm/functions.py

@@ -19,15 +19,13 @@ def func(x):
     return -result
 
 
-def ackley(x):
-    a = 20
-    b = 0.2
-    c = 2 * np.pi
+def ackley(x, a=20, b=0.2, c=2*np.pi):
 
-    first_sum = reduce(lambda acc, x: acc + np.power(x, 2), x, 0.)
-    second_sum = reduce(lambda acc, x: np.cos(c * x), x, 0.)
-
-    return -a * np.exp(-b * np.sqrt(1/x.size * first_sum)-np.exp(1/x.size * second_sum)) + a + np.exp(1)
+    x = np.asarray_chkfinite(x)  # ValueError if any NaN or Inf
+    n = len(x)
+    s1 = np.sum(x**2, axis=0)
+    s2 = np.sum(np.cos(c * x), axis=0)
+    return -a*np.exp(-b*np.sqrt(s1 / n)) - np.exp(s2 / n) + a + np.exp(1)
 
 
 FUNCTIONS = {