Merge pull request #545 from Tarun-Sreepada/main

found bug. fixed. hopefully

PAMI/partialPeriodicFrequentPattern/basic/GPFgrowth.py

@@ -306,9 +306,8 @@ def _getPerSup(self, arr):
         :type arr: array
         :return: locs
         """
-        arr = list(arr)
-        arr.append(self._maxTS)
-        arr.append(0)
+        arr = list(arr) + [self._maxTS, 0]
+        arr = list(set(arr))
         arr = np.sort(arr)
         arr = np.diff(arr)
 
@@ -342,11 +341,12 @@ def _construct(self, items, data):
         """
 
 
-        items = {k: v for k, v in items.items() if len(v) >= self._partialPeriodicPatterns__minSup and self._ratioCalc(v) >= self._partialPeriodicPatterns__minPR}
+        items = {k: v for k, v in items.items() if len(v) >= self._partialPeriodicPatterns__minSup}
 
-        #tested ok
         for item, ts in items.items():
-            self._partialPeriodicPatterns__finalPatterns[tuple([item])] = [len(ts), self._ratioCalc(ts)]
+            ratio = self._ratioCalc(ts)
+            if ratio >= self._partialPeriodicPatterns__minPR:
+                self._partialPeriodicPatterns__finalPatterns[tuple([item])] = [len(ts), ratio]
 
         root = _Node([], None, None)
         itemNodes = {}
@@ -364,7 +364,6 @@ def _construct(self, items, data):
 
         return root, itemNodes
 
-
     def _recursive(self, root, itemNode):
         """
         This method recursively constructs a pattern tree from the given root node,
@@ -410,14 +409,13 @@ def _recursive(self, root, itemNode):
             # maxPerResults = {item: self._getMaxPer(itemLocs[item], maxTS) for item in itemLocs if len(itemLocs[item]) >= minSup}
             maxPerResults = {item: self._ratioCalc(itemLocs[item]) for item in itemLocs if len(itemLocs[item]) >= self._partialPeriodicPatterns__minSup}
 
+            for item in maxPerResults:
+                if maxPerResults[item] >= self._partialPeriodicPatterns__minPR:
+                    self._partialPeriodicPatterns__finalPatterns[tuple(newRoot.item + [item])] = [len(itemLocs[item]), maxPerResults[item]]
 
             # Filter itemLocs based on minSup and maxPer
-            itemLocs = {k: len(v) for k, v in itemLocs.items() if k in maxPerResults and maxPerResults[k] >= self._partialPeriodicPatterns__minPR}
+            itemLocs = {k: len(v) for k, v in itemLocs.items() if len(v) >= self._partialPeriodicPatterns__minSup}
 
-            # Iterate over filtered itemLocs
-            for item in itemLocs:
-                self._partialPeriodicPatterns__finalPatterns[tuple(newRoot.item + [item])] = [itemLocs[item], maxPerResults[item]]
-
             if not itemLocs:
                 continue
 
@@ -437,14 +435,11 @@ def _recursive(self, root, itemNode):
 
             self._recursive(newRoot, newItemNodes)
 
-
     def mine(self):
-        self.__inputFile = self._partialPeriodicPatterns__iFile
         self._partialPeriodicPatterns__startTime = time.time()
         self._partialPeriodicPatterns__finalPatterns = {}
         self.__creatingItemSets()
 
-        # self.minPR = self.convert(self.minPR)
         self._partialPeriodicPatterns__minPR = float(self._partialPeriodicPatterns__minPR)
 
         self._maxTS = 0
@@ -581,13 +576,5 @@ def printResults(self):
         print("Total Memory in RSS", ap.getMemoryRSS())
         print("Total ExecutionTime in ms:", ap.getRuntime())
     else:
-        for i in [350]:
-            _ap = GPFgrowth('/Users/tarunsreepada/Downloads/Temporal_T10I4D100K.csv', i, 300, 0.7, '\t')
-            _ap.mine()
-            print("Total number of Maximal Partial Periodic Patterns:", len(_ap.getPatterns()))
-            _ap.save('/Users/tarunsreepada/Downloads/output1.txt')
-            print(_ap.getPatternsAsDataFrame())
-            print("Total Memory in USS:", _ap.getMemoryUSS())
-            print("Total Memory in RSS", _ap.getMemoryRSS())
-            print("Total ExecutionTime in ms:", _ap.getRuntime())
-        print("Error! The number of input parameters do not match the total number of parameters provided")
+
+        print("Error! The number of input parameters do not match the total number of parameters provided")

PAMI/partialPeriodicFrequentPattern/basic/PPF_DFS.py

@@ -63,6 +63,7 @@
 import numpy as np
 import pandas as pd
 
+
 class PPF_DFS(partialPeriodicPatterns):
     """
     **About this algorithm**
@@ -286,6 +287,7 @@ def _getPerSup(self, arr):
         arr = list(arr)
         arr.append(self._maxTS)
         arr.append(0)
+        arr = list(set(arr))
         arr = np.sort(arr)
         arr = np.diff(arr)
 
@@ -314,12 +316,12 @@ def __recursive(self, cands, items):
                 if len(intersection) >= self._partialPeriodicPatterns__minSup:
                     perSup = self._getPerSup(intersection)
                     ratio = perSup / (len(intersection) + 1)
+                    nCand = cands[i] + tuple([cands[j][-1]])
+                    newCands.append(nCand)
+                    nitems[nCand] = intersection
                     if ratio >= self._partialPeriodicPatterns__minPR:
-                        nCand = cands[i] + tuple([cands[j][-1]])
-                        newCands.append(nCand)
-                        nitems[nCand] = intersection
                         self._partialPeriodicPatterns__finalPatterns[nCand] = [len(intersection), ratio]
-            if len(newCands) > 1:
+            if len(newCands):
                 self.__recursive(newCands, nitems)
 
 
@@ -473,15 +475,5 @@ def printResults(self):
         print("Total Memory in RSS", ap.getMemoryRSS())
         print("Total ExecutionTime in ms:", ap.getRuntime())
     else:
-        for i in [350]:
-            #385
-            _ap = PPF_DFS('/Users/tarunsreepada/Downloads/Temporal_T10I4D100K.csv', i, 300, 0.7, '\t')
-            _ap.mine()
-            _ap.save('/Users/tarunsreepada/Downloads/output2.txt')
-            print(_ap.getPatternsAsDataFrame())
-            print("Total Memory in USS:", _ap.getMemoryUSS())
-            print("Total Memory in RSS", _ap.getMemoryRSS())
-            print("Total ExecutionTime in ms:", _ap.getRuntime())
         print("Error! The number of input parameters do not match the total number of parameters provided")
 
-