V1

Author: mohsystem

src/AlgorithmsTestingV3.py

@@ -4,10 +4,12 @@
  Algorithm testing .
 '''
 
+import time
+
 from src.KmeansAlgorithmV3 import KmeansAlgorithm
 from src.Utility import Utility
 
-import time
+
 class AlgorithmsTesting:
  def testKmeansPureImpl(self):
  '''
@@ -24,7 +26,7 @@ def testKmeansPureImpl(self):
  # needDimReduction = True
 
  bestCostFunction = 999999999
- bestNumberOfClusters =99999999
+ bestNumberOfClusters = 99999999
  bestWorkingCopyDS = None
  bestClusterMemberships = None
  bestClusterCenters = None
@@ -33,15 +35,14 @@ def testKmeansPureImpl(self):
  # Describe the data
  utility.describeDS(originalDS)
 
-
  kmeansAlgorithm = KmeansAlgorithm()
  # Call the K-mean function
  maxNumberOfClusters = 15
  prevoiusBestRunCostFunction = 999999999
  t = time.localtime()
- for numberOfClusters in range(2,maxNumberOfClusters):
+ for numberOfClusters in range(2, maxNumberOfClusters):
  workingCopyDS, clusterMemberships, clusterCenters, algorithmConvergeHistory, \
- costFunction= \
+ costFunction = \
  kmeansAlgorithm.kmeansAlgorithmPureImpl(originalDS=originalDS,
  reclassifcationIterationLimit=50,
  stopLimit=0.01,
@@ -58,11 +59,11 @@ def testKmeansPureImpl(self):
  bestCostFunction = costFunction
 
  # If no significant improvement, then break
- if prevoiusBestRunCostFunction - costFunction <=0.02 :
+ if prevoiusBestRunCostFunction - costFunction <= 0.02:
  break
  prevoiusBestRunCostFunction = costFunction
 
- print("Time in execution: " , (time.localtime().tm_sec - t.tm_sec) )
+ print("Time in execution: ", (time.localtime().tm_sec - t.tm_sec))
  # Draw the data points along with the clusters centers )
  utility.plot(bestClusterCenters, bestClusterMemberships, bestWorkingCopyDS,
  clusterBasedAlgorithmConvergeHistory)

src/KmeansAlgorithmV3.py

@@ -126,7 +126,7 @@ def kmeansAlgorithmPureImpl(_self, originalDS,
 
  # points assignments on each cluster result
 
-  ## print("Points assignments on each cluster result for iteration # ", currentIteration,
+ ## print("Points assignments on each cluster result for iteration # ", currentIteration,
  ## ": ", clusterMemberships)
  # print("Points assignments on each cluster result for iteration # ",currentIteration,
  # ": ", clusterMemberships1)
@@ -221,14 +221,13 @@ def kmeansAlgorithmPureImpl(_self, originalDS,
 
  costFunction = sumOfCostFunction / numberOfAssignedClusterPoints
 
-
  if costFunction <= stopLimit:
  isClusterCenterGetConverged = True
 
  ## print("algorithmConvergeHistory: ", algorithmConvergeHistory)
  algorithmConvergeHistory.append(costFunction)
  return workingCopyDS, clusterMemberships, clusterCenters, \
- algorithmConvergeHistory,costFunction
+ algorithmConvergeHistory, costFunction
 
 
 def euclideanDistance(currentClusterCenter, currentPoint):

src/Utility.py

@@ -48,7 +48,7 @@ def plot(self, clusterCenters, clusterMemberships, workingCopyDS, algorithmConve
  plt.tight_layout()
  plt.show()
 
- sns.lineplot(x=range(0+1, len(algorithmConvergeHistory)+1), y=algorithmConvergeHistory, marker='x')
+ sns.lineplot(x=range(0 + 1, len(algorithmConvergeHistory) + 1), y=algorithmConvergeHistory, marker='x')
  plt.xlabel('X')
  plt.ylabel('Y')
  plt.show()