MCPcopy Create free account
hub / github.com/Jack-Cherish/Machine-Learning / createTree

Function createTree

Regression Trees/regTrees.py:150–178  ·  view source on GitHub ↗

函数说明:树构建函数 Parameters: dataSet - 数据集合 leafType - 建立叶结点的函数 errType - 误差计算函数 ops - 包含树构建所有其他参数的元组 Returns: retTree - 构建的回归树 Website: http://www.cuijiahua.com/ Modify: 2017-12-12

(dataSet, leafType = regLeaf, errType = regErr, ops = (1, 4))

Source from the content-addressed store, hash-verified

148 return bestIndex, bestValue
149
150def createTree(dataSet, leafType = regLeaf, errType = regErr, ops = (1, 4)):
151 """
152 函数说明:树构建函数
153 Parameters:
154 dataSet - 数据集合
155 leafType - 建立叶结点的函数
156 errType - 误差计算函数
157 ops - 包含树构建所有其他参数的元组
158 Returns:
159 retTree - 构建的回归树
160 Website:
161 http://www.cuijiahua.com/
162 Modify:
163 2017-12-12
164 """
165 #选择最佳切分特征和特征值
166 feat, val = chooseBestSplit(dataSet, leafType, errType, ops)
167 #r如果没有特征,则返回特征值
168 if feat == None: return val
169 #回归树
170 retTree = {}
171 retTree['spInd'] = feat
172 retTree['spVal'] = val
173 #分成左数据集和右数据集
174 lSet, rSet = binSplitDataSet(dataSet, feat, val)
175 #创建左子树和右子树
176 retTree['left'] = createTree(lSet, leafType, errType, ops)
177 retTree['right'] = createTree(rSet, leafType, errType, ops)
178 return retTree
179
180def isTree(obj):
181 """

Callers 1

regTrees.pyFile · 0.70

Calls 2

chooseBestSplitFunction · 0.85
binSplitDataSetFunction · 0.85

Tested by

no test coverage detected