-
Notifications
You must be signed in to change notification settings - Fork 0
/
bTree.py
167 lines (146 loc) · 5.44 KB
/
bTree.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
import pickle
import random
# from common.helper import *
from benchmarkOption import *
# For dummy feature
FEATURE_USED ={
"linnerud": [0, 1, 2],
"cancer":[0,1,2],
"wine":[0,1,2],
"digits-10":[0,1,2],
"digits-12":[0,1,2],
"digits-15":[0,1,2],
"diabets-18":[0,1,2],
"diabets-20":[0,1,2]
}
class Node:
def __init__(self,pNode,threshold,leafValue,feature):
self.left = None
self.right = None
self.isLeaf = False
self.classId=None
self.height=0
self.featureIdx = feature
self.parent = pNode
if self.parent is not None:
self.height = self.parent.getHeight()+1
self.threshold = threshold
self.condition = 1 #Denotes lessThan operation or equal operation
self.leafValue = leafValue
def setAsLeaf(self):
self.isLeaf = True
for i, v in enumerate(self.leafValue):
if v != 0:
self.classId = i
break
def getHeight(self):
return self.height
# print(clf.tree_.threshold)
# print(clf.tree_.value)
# totalNum = 2**(clf.get_depth()+1) -1
class BinaryTree:
def __init__(self, name):
clf=None
self.modelName = name
with open("./model/"+name+".model", 'rb') as fid:
clf= pickle.load(fid)
self.thresholds = clf.tree_.threshold
self.values = clf.tree_.value
self.features = clf.tree_.feature
self.dummyLeafCnt = 0
# print("binary tree features are: ", self.features)
# self.featureUsed = clf.n_features_in_int
# print("nodes num is: ",len(self.thresholds))
# print(self.values)
# count used feature number here
usedFeatures={}
cnt=0
for f in self.features:
if f != -2:
if str(f) not in usedFeatures.keys():
usedFeatures[str(f)] = cnt
cnt+=1
self.usedFeatureLen = len(usedFeatures.keys())
# print("#Used_features is: ", self.usedFeatureLen)
# print(usedFeatures)
self.usedFeaturesMapping = usedFeatures.copy()
renamedFeatures = []
# self.features.copy()
for v in self.features:
if v== -2:
renamedFeatures.append(-2)
else:
renamedFeatures.append(usedFeatures[str(v)])
self.features = renamedFeatures
# print(self.features)
# print(clf.tree_.feature)
self.maxHeight = clf.get_depth()
self.dummyNodesCnt = 2**(self.maxHeight+1)-1-len(self.thresholds)
# nodesNum = len(thresholds)
# print(totalNum)
idx,self.root = self.middleOrderCreate(None,0)
def getMaxHeight(self):
return self.maxHeight
def getUsedFeatures(self):
return self.usedFeatureLen
def getFeatureIdxMapping(self):
return self.usedFeaturesMapping
def middleOrderCreate(self,parent,idx):
newIdx, middleNode = 0,None
leafValue = self.values[idx][0]
leafValue = [int(v) for v in leafValue]
# print("leafValue is: ",leafValue)
threshold = self.thresholds[idx]
if threshold == -2:
rndIndex = random.choice(FEATURE_USED[self.modelName])
newIdx, middleNode = idx + \
1, Node(parent, threshold, leafValue, rndIndex)
# print("leafValue: ",leafValue)
# self.dummyNodesCnt -= 1
self.addDummyNode(middleNode,leafValue)
else:
newIdx, middleNode = idx + \
1, Node(parent, threshold, leafValue, self.features[idx])
newIdx,leftNode = self.middleOrderCreate(middleNode,newIdx)
newIdx,rightNode = self.middleOrderCreate(middleNode,newIdx)
middleNode.left =leftNode
middleNode.right =rightNode
return newIdx,middleNode
def addDummyNode(self,parent,leafValue):
# self.dummyNodesCnt += 1
if parent.getHeight() < self.getMaxHeight(): # Automatically insert dummy nodes with nonleaf node
leftNode = Node(parent, -2, leafValue,
random.choice(FEATURE_USED[self.modelName]))
rightNode = Node(parent, -2, leafValue,
random.choice(FEATURE_USED[self.modelName]))
parent.left = leftNode
parent.right = rightNode
self.addDummyNode(leftNode,leafValue)
self.addDummyNode(rightNode,leafValue)
else:
# self.dummyNodesCnt -= 1
# self.dummyLeafCnt += 1
parent.setAsLeaf()
def getNodesInfo(self,CONVERT_FACTOR):
# print(self.modelName," is with ",self.dummyNodesCnt," dummy nodes.")
# print(self.modelName, " is with ", self.maxHeight, " depth.")
queue = [self.root]
nonLeafNodes=[]
leafNodes=[]
while len(queue) > 0:
top = queue.pop(0)
if top.isLeaf:
leafNodes.append(top.classId)
else:
queue.append(top.left)
queue.append(top.right)
t = int(top.threshold*CONVERT_FACTOR)
# if t == -200:
# t=1
tuple = (top.featureIdx,t )
nonLeafNodes.append(tuple)
return leafNodes, nonLeafNodes
# print("nonleaf len is: ", len(nonLeafNodes))
# print("leaf len is: ", len(leafNodes))
# print(leafNodes)
# print(nonLeafNodes)