1. import matplotlib.pyplot as plt
  2.  
  3. decisionNode = dict(boxstyle="sawtooth", fc="0.8")
  4. leafNode = dict(boxstyle="round4", fc="0.8")
  5. arrow_args = dict(arrowstyle="<-")
  6.  
  7. def getNumLeafs(myTree):
  8.     numLeafs = 0
  9.     for i in myTree.keys():
  10.         firstStr = i
  11.         break
  12.     secondDict = myTree[firstStr]
  13.     for key in secondDict.keys():
  14.         if type(secondDict[key]).__name__=='dict':
  15.             numLeafs += getNumLeafs(secondDict[key])
  16.         else:   numLeafs +=1
  17.     return numLeafs
  18.  
  19. def getTreeDepth(myTree):
  20.     maxDepth = 0
  21.     for i in myTree.keys():
  22.         firstStr = i
  23.         break
  24.     secondDict = myTree[firstStr]
  25.     for key in secondDict.keys():
  26.         if type(secondDict[key]).__name__=='dict':
  27.             thisDepth = 1 + getTreeDepth(secondDict[key])
  28.         else:   thisDepth = 1
  29.         if thisDepth > maxDepth: maxDepth = thisDepth
  30.     return maxDepth
  31.  
  32. def plotNode(nodeTxt, centerPt, parentPt, nodeType):
  33.     createPlot.ax1.annotate(nodeTxt, xy=parentPt,  xycoords='axes fraction',xytext=centerPt, textcoords='axes fraction',va="center", ha="center", bbox=nodeType, arrowprops=arrow_args )
  34.  
  35. def plotMidText(cntrPt, parentPt, txtString):
  36.     xMid = (parentPt[0]-cntrPt[0])/2.0 + cntrPt[0]
  37.     yMid = (parentPt[1]-cntrPt[1])/2.0 + cntrPt[1]
  38.     createPlot.ax1.text(xMid, yMid, txtString, va="center", ha="center", rotation=30)
  39.  
  40. def plotTree(myTree, parentPt, nodeTxt):
  41.     numLeafs = getNumLeafs(myTree)
  42.     depth = getTreeDepth(myTree)
  43.     for i in myTree.keys():
  44.         firstStr = i
  45.         break
  46.     cntrPt = (plotTree.xOff + (1.0 + float(numLeafs))/2.0/plotTree.totalW, plotTree.yOff)
  47.     plotMidText(cntrPt, parentPt, nodeTxt)
  48.     plotNode(firstStr, cntrPt, parentPt, decisionNode)
  49.     secondDict = myTree[firstStr]
  50.     plotTree.yOff = plotTree.yOff - 1.0/plotTree.totalD
  51.     for key in secondDict.keys():
  52.         if type(secondDict[key]).__name__=='dict':
  53.             plotTree(secondDict[key],cntrPt,str(key))
  54.         else:
  55.             plotTree.xOff = plotTree.xOff + 1.0/plotTree.totalW
  56.             plotNode(secondDict[key], (plotTree.xOff, plotTree.yOff), cntrPt, leafNode)
  57.             plotMidText((plotTree.xOff, plotTree.yOff), cntrPt, str(key))
  58.     plotTree.yOff = plotTree.yOff + 1.0/plotTree.totalD
  59.  
  60. def createPlot(inTree):
  61.     fig = plt.figure(1, facecolor='white')
  62.     fig.clf()
  63.     axprops = dict(xticks=[], yticks=[])
  64.     createPlot.ax1 = plt.subplot(111, frameon=False, **axprops)
  65.     #createPlot.ax1 = plt.subplot(111, frameon=False) #ticks for demo puropses
  66.     plotTree.totalW = float(getNumLeafs(inTree))
  67.     plotTree.totalD = float(getTreeDepth(inTree))
  68.     plotTree.xOff = -0.5/plotTree.totalW; plotTree.yOff = 1.0;
  69.     plotTree(inTree, (0.5,1.0), '')
  70.     plt.show()
  71.  
  72. def retrieveTree(i):
  73.     listOfTrees =[{'no surfacing': {0: 'no', 1: {'flippers': {0: 'no', 1: 'yes'}}}},
  74.                   {'no surfacing': {0: 'no', 1: {'flippers': {0: {'head': {0: 'no', 1: 'yes'}}, 1: 'no'}}}}
  75.                   ]
  76.     return listOfTrees[i]
  77.  
  78. thisTree = retrieveTree(0)
  79. createPlot(thisTree)
  80. thisTree = retrieveTree(1)
  81. createPlot(thisTree)

  1. import numpy as np
  2. import operator as op
  3. from math import log
  4.  
  5. def calcShannonEnt(dataSet):
  6.     labelCounts = {}
  7.     for featVec in dataSet:
  8.         currentLabel = featVec[-1]
  9.         if(currentLabel not in labelCounts.keys()):
  10.             labelCounts[currentLabel] = 0
  11.         labelCounts[currentLabel] += 1
  12.     shannonEnt = 0.0
  13.     rowNum = len(dataSet)
  14.     for key in labelCounts:
  15.         prob = float(labelCounts[key])/rowNum
  16.         shannonEnt -= prob * log(prob,2)
  17.     return shannonEnt
  18.  
  19. def splitDataSet(dataSet, axis, value):
  20.     retDataSet = []
  21.     for featVec in dataSet:
  22.         if(featVec[axis] == value):
  23.             reducedFeatVec = featVec[:axis]
  24.             reducedFeatVec.extend(featVec[axis+1:])
  25.             retDataSet.append(reducedFeatVec)
  26.     return retDataSet
  27.  
  28. def chooseBestFeatureToSplit(dataSet):
  29.     numFeatures = np.shape(dataSet)[1]-1
  30.     baseEntropy = calcShannonEnt(dataSet)
  31.     bestInfoGain = 0.0
  32.     bestFeature = -1
  33.     for i in range(numFeatures):
  34.         featList = [example[i] for example in dataSet]
  35.         uniqueVals = set(featList)
  36.         newEntropy = 0.0
  37.         for value in uniqueVals:
  38.             subDataSet = splitDataSet(dataSet, i, value)
  39.             prob = len(subDataSet)/float(len(dataSet))
  40.             newEntropy += prob * calcShannonEnt(subDataSet)
  41.         infoGain = baseEntropy - newEntropy
  42.         if (infoGain > bestInfoGain):
  43.             bestInfoGain = infoGain
  44.             bestFeature = i
  45.     return bestFeature 
  46.  
  47. def majorityCnt(classList):
  48.     classCount={}
  49.     for vote in classList:
  50.         if(vote not in classCount.keys()):
  51.             classCount[vote] = 0
  52.         classCount[vote] += 1
  53.     sortedClassCount = sorted(classCount.items(), key=op.itemgetter(1), reverse=True)
  54.     return sortedClassCount[0][0]
  55.  
  56. def createTree(dataSet,labels):
  57.     classList = [example[-1] for example in dataSet]
  58.     if(classList.count(classList[0]) == len(classList)):
  59.         return classList[0]
  60.     if len(dataSet[0]) == 1:
  61.         return majorityCnt(classList)
  62.     bestFeat = chooseBestFeatureToSplit(dataSet)
  63.     bestFeatLabel = labels[bestFeat]
  64.     myTree = {bestFeatLabel:{}}
  65.     del(labels[bestFeat])
  66.     featValues = [example[bestFeat] for example in dataSet]
  67.     uniqueVals = set(featValues)
  68.     for value in uniqueVals:
  69.         subLabels = labels[:]
  70.         myTree[bestFeatLabel][value] = createTree(splitDataSet(dataSet, bestFeat, value),subLabels)
  71.     return myTree
  72.  
  73. def classify(inputTree,featLabels,testVec):
  74.     for i in inputTree.keys():
  75.         firstStr = i
  76.         break
  77.     secondDict = inputTree[firstStr]
  78.     featIndex = featLabels.index(firstStr)
  79.     key = testVec[featIndex]
  80.     valueOfFeat = secondDict[key]
  81.     if isinstance(valueOfFeat, dict):
  82.         classLabel = classify(valueOfFeat, featLabels, testVec)
  83.     else:
  84.         classLabel = valueOfFeat
  85.     return classLabel
  86.  
  87. data = open("D:\\LearningResource\\machinelearninginaction\\Ch03\\lenses.txt")
  88. dataSet = [inst.strip().split("\t") for inst in data.readlines()]
  89. print(dataSet)
  90. print(np.shape(dataSet))
  91. labels = ["age","prescript","astigmatic","tearRate"]
  92. tree = createTree(dataSet,labels)
  93. print(tree)
  94.  
  95. import matplotlib.pyplot as plt
  96.  
  97. decisionNode = dict(boxstyle="sawtooth", fc="0.8")
  98. leafNode = dict(boxstyle="round4", fc="0.8")
  99. arrow_args = dict(arrowstyle="<-")
  100.  
  101. def getNumLeafs(myTree):
  102.     numLeafs = 0
  103.     for i in myTree.keys():
  104.         firstStr = i
  105.         break
  106.     secondDict = myTree[firstStr]
  107.     for key in secondDict.keys():
  108.         if type(secondDict[key]).__name__=='dict':
  109.             numLeafs += getNumLeafs(secondDict[key])
  110.         else:   numLeafs +=1
  111.     return numLeafs
  112.  
  113. def getTreeDepth(myTree):
  114.     maxDepth = 0
  115.     for i in myTree.keys():
  116.         firstStr = i
  117.         break
  118.     secondDict = myTree[firstStr]
  119.     for key in secondDict.keys():
  120.         if type(secondDict[key]).__name__=='dict':
  121.             thisDepth = 1 + getTreeDepth(secondDict[key])
  122.         else:   thisDepth = 1
  123.         if thisDepth > maxDepth: maxDepth = thisDepth
  124.     return maxDepth
  125.  
  126. def plotNode(nodeTxt, centerPt, parentPt, nodeType):
  127.     createPlot.ax1.annotate(nodeTxt, xy=parentPt,  xycoords='axes fraction',xytext=centerPt, textcoords='axes fraction',va="center", ha="center", bbox=nodeType, arrowprops=arrow_args )
  128.  
  129. def plotMidText(cntrPt, parentPt, txtString):
  130.     xMid = (parentPt[0]-cntrPt[0])/2.0 + cntrPt[0]
  131.     yMid = (parentPt[1]-cntrPt[1])/2.0 + cntrPt[1]
  132.     createPlot.ax1.text(xMid, yMid, txtString, va="center", ha="center", rotation=30)
  133.  
  134. def plotTree(myTree, parentPt, nodeTxt):
  135.     numLeafs = getNumLeafs(myTree)
  136.     depth = getTreeDepth(myTree)
  137.     for i in myTree.keys():
  138.         firstStr = i
  139.         break
  140.     cntrPt = (plotTree.xOff + (1.0 + float(numLeafs))/2.0/plotTree.totalW, plotTree.yOff)
  141.     plotMidText(cntrPt, parentPt, nodeTxt)
  142.     plotNode(firstStr, cntrPt, parentPt, decisionNode)
  143.     secondDict = myTree[firstStr]
  144.     plotTree.yOff = plotTree.yOff - 1.0/plotTree.totalD
  145.     for key in secondDict.keys():
  146.         if type(secondDict[key]).__name__=='dict':
  147.             plotTree(secondDict[key],cntrPt,str(key))
  148.         else:
  149.             plotTree.xOff = plotTree.xOff + 1.0/plotTree.totalW
  150.             plotNode(secondDict[key], (plotTree.xOff, plotTree.yOff), cntrPt, leafNode)
  151.             plotMidText((plotTree.xOff, plotTree.yOff), cntrPt, str(key))
  152.     plotTree.yOff = plotTree.yOff + 1.0/plotTree.totalD
  153.  
  154. def createPlot(inTree):
  155.     fig = plt.figure(1, facecolor='white')
  156.     fig.clf()
  157.     axprops = dict(xticks=[], yticks=[])
  158.     createPlot.ax1 = plt.subplot(111, frameon=False, **axprops)
  159.     #createPlot.ax1 = plt.subplot(111, frameon=False) #ticks for demo puropses
  160.     plotTree.totalW = float(getNumLeafs(inTree))
  161.     plotTree.totalD = float(getTreeDepth(inTree))
  162.     plotTree.xOff = -0.5/plotTree.totalW; plotTree.yOff = 1.0;
  163.     plotTree(inTree, (0.5,1.0), '')
  164.     plt.show()
  165.  
  166. createPlot(tree)

吴裕雄 python 机器学习-DMT（2）的更多相关文章

  1. 吴裕雄 python 机器学习-DMT（1）

    import numpy as np import operator as op from math import log def createDataSet(): dataSet = [[1, 1, ...

  2. 吴裕雄 python 机器学习——分类决策树模型

    import numpy as np import matplotlib.pyplot as plt from sklearn import datasets from sklearn.model_s ...

  3. 吴裕雄 python 机器学习——回归决策树模型

    import numpy as np import matplotlib.pyplot as plt from sklearn import datasets from sklearn.model_s ...

  4. 吴裕雄 python 机器学习——线性判断分析LinearDiscriminantAnalysis

    import numpy as np import matplotlib.pyplot as plt from matplotlib import cm from mpl_toolkits.mplot ...

  5. 吴裕雄 python 机器学习——逻辑回归

    import numpy as np import matplotlib.pyplot as plt from matplotlib import cm from mpl_toolkits.mplot ...

  6. 吴裕雄 python 机器学习——ElasticNet回归

    import numpy as np import matplotlib.pyplot as plt from matplotlib import cm from mpl_toolkits.mplot ...

  7. 吴裕雄 python 机器学习——Lasso回归

    import numpy as np import matplotlib.pyplot as plt from sklearn import datasets, linear_model from s ...

  8. 吴裕雄 python 机器学习——岭回归

    import numpy as np import matplotlib.pyplot as plt from sklearn import datasets, linear_model from s ...

  9. 吴裕雄 python 机器学习——线性回归模型

    import numpy as np from sklearn import datasets,linear_model from sklearn.model_selection import tra ...

随机推荐

  1. 怎样找到微信小程序功能呢？

    怎样找到微信小程序功能呢? 17年1月9号这个功能刚推出时,非常火,但我在微信中迟迟找不到微信小程序功能,微信明明更新到了最新版也没有:后来经过以下步骤才找到这个功能! 1.微信版本不能过低,这个一般 ...

  2. Django-models的字段类型

    model的field类型 1.models.AutoField   ---自增列 = int(11)    如果没有的话,默认会生成一个名称为 id 的列,如果要显示的自定义一个自增列,必须将给列设 ...

  3. useful tools and website

    1.https://www.processon.com/   在线流程图制作网站 2.http://www.easyicon.net/  专门下载图标的网站 3.https://www.lfd.uci ...

  4. CPU Rings, Privilege, and Protection.CPU的运行环, 特权级与保护

    原文标题:CPU Rings, Privilege, and Protection 原文地址:http://duartes.org/gustavo/blog/ [注:本人水平有限,只好挑一些国外高手的 ...

  5. Spring之jdbcTemplate：查询的三种方式（单个值、单个对象、对象集合）

    JdbcTemplateDemo2.java package helloworld.jdbcTemplate; import org.springframework.jdbc.core.JdbcTem ...

  6. WPF开发ArcGis系统时的异常信息: ArcGIS product not specified. You must first bind to an ArcGIS version prior to using any ArcGIS components.

    “System.Runtime.InteropServices.COMException”类型的未经处理的异常在 Arcgis_Test.exe 中发生 其他信息: ArcGIS product no ...

  7. [Unity动画]05.Entry & Exit & Any State

    0.状态机如下: Any State->Dying:isDying为true Dying->Reviving:isDying为false Reviving->Exit:isDying ...

  8. 【每日一学】pandas_透视表函数&交叉表函数

    每日一悟 [分开工作内外8小时] 前一个月,我经常把工作内的问题带到路上.地铁上.睡觉前,甚至是周末. 然而很快发现,我工作外的成就几乎没有,而工作内的进展也并不理想. 仔细想想,工作外是需要学新东西 ...

  9. 【Social listening实操】从社交媒体传播和文本挖掘角度解读《欢乐颂2》

    本文转自知乎 作者:苏格兰折耳喵 ----------------------------------------------------- 作为数据分析爱好者,本文作者将想从数据的角度去解读< ...

  10. Android自定义View学习笔记（一）

    绘制基础 参考:HenCoder Android 开发进阶: 自定义 View 1-1 绘制基础 Paint详解 参考:HenCoder Android 开发进阶: 自定义 View 1-2 Pain ...