/
model_index.py
224 lines (174 loc) · 8.56 KB
/
model_index.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
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
from math import exp,log
from sys import stdout
from dendropy.treecalc import false_positives_and_negatives, euclidean_distance
##PUT ALL PUBLICS UP TOP
class model_index:
def __init__(self,args):
self.models={}
self.ranked_models=[]
self.use_dtc=args.dtc
self.use_cong=args.cong
self.clist=args.clist
self.min_criterion=self.__create_crit_dict()
self.sumweights=self.__create_crit_dict()
if self.use_dtc==True:
self.min_criterion['DTC']={}
self.sumweights['DTC']={}
for cfactor in self.clist:
self.min_criterion['DTC'][cfactor]=0.0
self.sumweights['DTC'][cfactor]=0.0
self.dtc_sum_dict={}
def __create_crit_dict(self):
dict={}
dict['AIC']={}
dict['AIC']['raw']=0.0
dict['BIC']={}
for cfactor in self.clist:
dict['AIC'][cfactor]= 0.0
dict['BIC'][cfactor]= 0.0
return dict
def add_model(self,model):
self.models[model.name]=model
##helper for calc dtcs
def calc_dtcs(self):
for cfactor in self.clist:
min_bic=self.min_criterion['BIC'][cfactor]
for model in self.models:
dtc_sum=0.0
model_tree=self.models[model].tree
for other_model in self.models:
if other_model==model: continue
other_tree=self.models[other_model].tree
distance=euclidean_distance(model_tree,other_tree)
if distance <= 0.0: continue
other_bic=self.models[other_model].criterion['BIC'][cfactor]
power = log(distance) - other_bic + min_bic
if power > -30.0:
dtc_sum+=exp(power)
self.models[model].criterion['DTC'][cfactor]=dtc_sum
#helper for fill deltas
def __get_minima(self,metric):
for crit in self.min_criterion[metric]:
best_model = min(self.models.itervalues(),key=lambda x: x.criterion[metric][crit])
self.min_criterion[metric][crit]=best_model.criterion[metric][crit]
def calc_deltas(self,metric):
self.__get_minima(metric)
for crit in self.min_criterion[metric]:
min=self.min_criterion[metric][crit]
for model in self.models:
crit_of_model=self.models[model].criterion[metric][crit]
self.models[model].delta[metric][crit]=crit_of_model-min
###Helper for calc weights
def __fill_pre_weights_get_sum_weights(self,metric):
for crit in self.sumweights[metric]:
for model in self.models:
delta=self.models[model].delta[metric][crit]
#print model, "DELTA", delta
preweight=exp(-1.0*delta/2.0)
#print model, "PREWIGHT",preweight
self.models[model].preweights[metric][crit]=preweight
self.sumweights[metric][crit]+=preweight
def calc_weights(self,metric):
self.__fill_pre_weights_get_sum_weights(metric)
for crit in self.sumweights[metric]:
sumweight=self.sumweights[metric][crit]
#print "SUM WEIGHT", sumweight
for model in self.models:
preweight=self.models[model].preweights[metric][crit]
self.models[model].weights[metric][crit]=preweight/sumweight
def rank_criteria_by_weight(self,metric):
for crit in self.sumweights[metric]:
self.ranked_models=self.models.values()
self.ranked_models.sort(key=lambda model: model.weights[metric][crit],reverse=True)
rank_count=1
for model in self.ranked_models:
self.models[model.name].rank[metric][crit]=rank_count
rank_count+=1
def sort_models(self,metric,crit):
self.ranked_models=self.models.values()
self.ranked_models.sort(key=lambda x: x.weights[metric][crit],reverse=True)
def calc_congruence(self,num_best_models):
for model in self.ranked_models:
for best_index in range(num_best_models):
if self.ranked_models[best_index]==model:
model.cong_list.append(' * ')
else:
congruence=false_positives_and_negatives(reference_tree=self.ranked_models[best_index].tree,test_tree=model.tree)
model.cong_list.append(congruence)
def print_delta_table(self,metric,crit):
stdout.write('\nSelected statistics for '+metric+'-'+crit+':\n')
stdout.flush()
stdout.write('='*127+'\n')
stdout.write(" "*10+'\t')
stdout.write("{0:<20}\t".format('delta '+metric+'-'+crit))
stdout.write("{0:<20}\t".format(metric+'-'+crit))
stdout.write("{0:<20}\t".format(metric+'-'+crit+' weight'))
stdout.write('\n'+'='*127+'\n')
stdout.flush()
for model in self.ranked_models:
stdout.write("{0:<10}\t".format(model.model_name+' +'+model.param_names))
delta="{0:.6f}".format(model.delta[metric][crit])
criterion="{0:.6f}".format(model.criterion[metric][crit])
weight="{0:.6f}".format(model.weights[metric][crit])
stdout.write("{0:<20}\t{1:<20}\t{2:<20}\t\n".format(delta,criterion,weight))
stdout.flush()
#This probably needs reworking
def print_big_table(self,metric,crit):
stdout.write('\nWeight table sorted by '+metric+'-'+crit+':\n')
stdout.flush()
stdout.write('='*127+'\n')
stdout.write(" "*10+'\t')
for criterion in ['AIC-raw','AIC-length','AIC-shannon','AIC-c3','BIC-length','BIC-shannon','BIC-c3']:
stdout.write("{0:<15}\t".format(criterion))
stdout.write('\n'+'='*127+'\n')
stdout.flush()
for model in self.ranked_models:
stdout.write("{0:<8}\t".format(model.model_name+' +'+model.param_names))
#print AIC-raw first
weight="{0:.6f}".format(model.weights['AIC']['raw'])
rank=str(model.rank['AIC']['raw'])
rank="({0})".format(rank.zfill(2))
stdout.write("{weight:>8} {rank}\t".format(weight=weight,rank=rank))
for framework in ['AIC','BIC']:
for criterion in self.clist:
weight="{0:.6f}".format(model.weights[framework][criterion])
#jumping through hoops because I cant figure out how to get str.format
#to both truncate floats and justify
rank=str(model.rank[framework][criterion])
rank="({0})".format(rank.zfill(2))
stdout.write("{weight:<8} {rank}\t".format(weight=weight,rank=rank))
stdout.write('\n')
stdout.flush()
##helper for print big table
def print_optional_table(self,metric,crit,num_best_models):
if self.use_dtc==False and self.use_cong==False:
return
stdout.write('='*127+'\n')
stdout.write(" "*10+'\t')
if self.use_dtc==True:
for cfactor in self.clist:
criterion='DTC-'+cfactor
stdout.write("{0:<15}\t".format(criterion))
if self.use_cong==True:
stdout.write(' ')
for model in self.ranked_models[:num_best_models]:
stdout.write("{0:<10}\t".format(model.model_name+' +'+model.param_names+' SD'))
stdout.write('\n'+'='*127+'\n')
stdout.flush()
for model in self.ranked_models:
stdout.write("{0:<10}\t".format(model.model_name+' +'+model.param_names))
if self.use_dtc==True:
for criterion in self.clist:
weight="{0:.6f}".format(model.weights['DTC'][criterion])
#jumping through hoops because I cant figure out how to get str.format
#to both truncate floats and justify
rank=str(model.rank['DTC'][criterion])
rank="({0})".format(rank.zfill(2))
stdout.write("{weight:<8} {rank}\t".format(weight=weight,rank=rank))
if self.use_cong==True:
if self.use_dtc==True:
stdout.write('| ')
for i in range(num_best_models):
stdout.write("{cong:<10}\t".format(cong=model.cong_list[i]))
stdout.write('\n')
stdout.flush()