Python3《機器學習實戰》代碼筆記（十二）--- FP-growth算法

參考資料:

機器學習實戰

'''
@version: 0.0.1
@Author: tqrs
@dev: python3 vscode
@Date: 2019-11-12 09:29:53
@LastEditTime: 2019-11-12 12:24:30
@FilePath: \\機器學習實戰\\12-FP-growth算法\\FPGrowth.py
@Descripttion: 只需對數據庫進行兩次掃描,第一次對所有元素項出現次數進行統計;第二次只考慮頻繁元素,它能夠更爲高效的挖掘數據
'''


class treeNode:
    def __init__(self, nameValue, numOccur, parentNode):
        # 節點元素名稱
        self.name = nameValue
        # 出現次數
        self.count = numOccur
        # 指向下一個相似節點的指針
        self.nodeLink = None
        # 指向父節點的指針
        self.parent = parentNode
        # 指向子節點的字典，以子節點的元素名稱爲鍵，指向子節點的指針爲值
        self.children = {}

    # 增加節點的出現次數值
    def inc(self, numOccur):
        self.count += numOccur

    # 輸出節點和子節點的FP樹結構
    def disp(self, ind=1):
        print(' ' * ind, self.name, ' ', self.count)
        for child in self.children.values():
            child.disp(ind + 1)


def createTree(dataSet, minSup=1):
    """
    [summary]:將數據集轉化爲FP樹
    1. 遍歷數據集，統計各元素項出現次數，創建頭指針表
    2. 移除頭指針表中不滿足最小值尺度的元素項
    3. 第二次遍歷數據集，創建FP樹。對每個數據集中的項集：
        3.1 初始化空FP樹
        3.2 對每個項集進行過濾和重排序
        3.3 使用這個項集更新FP樹，從FP樹的根節點開始：
            3.3.1 如果當前項集的第一個元素項存在於FP樹當前節點的子節點中，則更新這個子節點的計數值
            3.3.2 否則，創建新的子節點，更新頭指針表
            3.3.3 對當前項集的其餘元素項和當前元素項的對應子節點遞歸3.3的過程
          
    Arguments:
        dataSet  -- 數據集
    
    Keyword Arguments:
        minSup {int} -- 最小支持度 (default: {1})
    
    Returns:
        [type] -- [description]
    """
    # 第一次遍歷數據集，創建頭指針表
    headerTable = {}
    for trans in dataSet:
        for item in trans:
            headerTable[item] = headerTable.get(item, 0) + dataSet[trans]
    # 移除不滿足最小支持度的元素項
    for k in list(headerTable.keys()):
        if headerTable[k] < minSup:
            del (headerTable[k])
    # 空元素集，返回空
    freqItemSet = set(headerTable.keys())
    if len(freqItemSet) == 0:
        return None, None
    # 增加一個數據項，用於存放指向相似元素項指針
    for k in headerTable:
        headerTable[k] = [headerTable[k], None]
    # 創造根節點
    retTree = treeNode('Null Set', 1, None)
    for tranSet, count in dataSet.items():
        # 對一個項集tranSet，記錄其中每個元素項的全局頻率，用於排序
        localD = {}
        for item in tranSet:
            if item in freqItemSet:
                localD[item] = headerTable[item][0]
        if len(localD) > 0:
            # 根據全局頻率對每個事務中的元素進行排序
            orderedItems = [
                v[0] for v in sorted(
                    localD.items(), key=lambda p: p[1], reverse=True)
            ]
            # 使用排序後的元素項集對樹進行填充
            updateTree(orderedItems, retTree, headerTable, count)
    return retTree, headerTable


def updateTree(items, inTree, headerTabel, count):
    # 有該元素項時計數值+1
    if items[0] in inTree.children:
        inTree.children[items[0]].inc(count)
    else:
        # 沒有這個元素項時創建一個新節點
        inTree.children[items[0]] = treeNode(items[0], count, inTree)
        if headerTabel[items[0]][1] == None:
            headerTabel[items[0]][1] = inTree.children[items[0]]
        else:
            updateHeader(headerTabel[items[0]][1], inTree.children[items[0]])
    # 對剩下的元素項迭代調用updateTree函數
    if len(items) > 1:
        updateTree(items[1::], inTree.children[items[0]], headerTabel, count)


def updateHeader(nodeToTest, targetNode):
    # 獲取頭指針表中該元素項對應的單鏈表的尾節點,然後將其指向新節點targetNode
    while (nodeToTest.nodeLink != None):
        nodeToTest = nodeToTest.nodeLink
    nodeToTest.nodeLink = targetNode


def loadSimpDat():
    # 加載數據集
    simpDat = [['r', 'z', 'h', 'j', 'p'],
               ['z', 'y', 'x', 'w', 'v', 'u', 't', 's'], ['z'],
               ['r', 'x', 'n', 'o', 's'], ['y', 'r', 'x', 'z', 'q', 't', 'p'],
               ['y', 'z', 'x', 'e', 'q', 's', 't', 'm']]
    return simpDat


def createInitSet(dataSet):
    # 生成數據集
    retDict = {}
    for trans in dataSet:
        retDict[frozenset(trans)] = 1
    return retDict


def test_fp():
    simpDat = loadSimpDat()
    initSet = createInitSet(simpDat)
    myFPtree, myHeaderTab = createTree(initSet, 3)
    myFPtree.disp()


def ascendTree(leafNode, prefixPath):
    # 迭代上溯整課樹
    if leafNode.parent != None:
        prefixPath.append(leafNode.name)
        ascendTree(leafNode.parent, prefixPath)


def findPrefixPath(basePat, treeNode):  # treeNode comes from header table
    # 創建前綴路徑
    condPats = {}
    while treeNode != None:
        prefixPath = []
        ascendTree(treeNode, prefixPath)
        if len(prefixPath) > 1:
            condPats[frozenset(prefixPath[1:])] = treeNode.count
        treeNode = treeNode.nodeLink
    return condPats


def test_pre():
    simpDat = loadSimpDat()
    initSet = createInitSet(simpDat)
    myFPtree, myHeaderTab = createTree(initSet, 3)
    condPats = findPrefixPath('r', myHeaderTab['r'][1])
    print(condPats)


def mineTree(inTree, headerTable, minSup, preFix, freqItemList):
    """
    [summary]:遞歸查找頻繁項集
    
    Arguments:
        inTree {[type]} -- [description]
        headerTable {[type]} -- [description]
        minSup {[type]} -- [description]
        preFix {[type]} -- [description]
        freqItemList  -- 頻繁項集列表
    """
    # 對頭指針表中的元素項按其出現頻率進行排序(默認從小到大)
    bigL = [v[0] for v in sorted(headerTable.items(), key=lambda p: p[1][0])]
    for basePat in bigL:  # start from bottom of header table
        newFreqSet = preFix.copy()
        newFreqSet.add(basePat)
        freqItemList.append(newFreqSet)
        condPattBases = findPrefixPath(basePat, headerTable[basePat][1])
        myCondTree, myHead = createTree(condPattBases, minSup)
        if myHead != None:
            print('conditional tree for: ', newFreqSet)
            myCondTree.disp(1)
            mineTree(myCondTree, myHead, minSup, newFreqSet, freqItemList)


def test_mineTree():
    simpDat = loadSimpDat()
    initSet = createInitSet(simpDat)
    myFPtree, myHeaderTab = createTree(initSet, 3)
    freqItems = []
    mineTree(myFPtree, myHeaderTab, 3, set([]), freqItems)


if __name__ == '__main__':
    paresdDat = [
        line.split()
        for line in open(r'.\12-FP-growth算法\kosarak.dat').readlines()
    ]

    initSet = createInitSet(paresdDat)
    myFPtree, myHeaderTab = createTree(initSet, 100000)
    myFreaList = []
    mineTree(myFPtree, myHeaderTab, 100000, set([]), myFreaList)
    print('len:', len(myFreaList), 'myFreaList', myFreaList)