-
Notifications
You must be signed in to change notification settings - Fork 0
/
_RUN_Me.R
275 lines (230 loc) · 10.9 KB
/
_RUN_Me.R
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
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
#####################################################################
#####################################################################
# load packages
#####################################################################
#####################################################################
library(SOAR) # to store data that will be used by all models
library(Matrix) # to handle sparse matrix
library(tau) # for n-grams
library(RTextTools) # to stem words
#####################################################################
#####################################################################
# set and store working directory
#####################################################################
#####################################################################
wd_path<-"~/Documents/asap3_short_answer"
setwd(wd_path)
Store(wd_path)
#####################################################################
#####################################################################
# set and store parameters to are used for all items
#####################################################################
#####################################################################
# training version
Version="V1"
Store(Version)
# item list
item_list=
c(
"3_51802_TS_51802_1","5_50786_TS_50786_1","5_50868_TB_50868_1","5_50932_TS_50932_1",
"5_53299_TB_53299_1","6_50577_TS_50577_1","6_52909_TB_52909_1","7_46115_TS_46115_1",
"7_46121_TS_46121_1","7_46282_TS_46282_1","7_46348_TS_46348_1","7_46450_TS_46450_1",
"7_46597_TS_46597_1","7_46793_TS_46793_1","8_46203_TB_46203_1","8_46517_TB_46517_1",
"8_48558_TS_48558_1","8_48560_TB_48560_1","8_53040_TB_53041_1","9_45535_TS_45535_1",
"9_48247_TB_48247_1","9_51416_TS_51416_1","10_46619_TS_46619_1","11_49151_TS_49151_1",
"11_49790_TS_49790_1")
Store(item_list)
# name of response to be modelled
Response_name="Final_Score"
Store(Response_name)
# nb of cores for parallel processing
nb_cores=5
# nb of folds for cross validation
nb_folds_training=10
Store(nb_cores,nb_folds_training)
source("_SETTING.R")
source("_VOCAB.R")
#####################################################################
#####################################################################
# inform if it is a TRAIN or TEST RUN
#####################################################################
#####################################################################
RUN_TYPE = "TRAIN"
Store(RUN_TYPE)
print("############################################################")
print(RUN_TYPE)
#####################################################################
for (item in item_list) {
print("############################################################")
print(paste("Item:",item))
print("############################################################")
Store(item)
print("############################################################")
print("# Read data")
#####################################################################
source("_READ_ONE_TRAIN_FILE.R")
print("############################################################")
print("# Generate Proxies")
print("############################################################")
source("_PROXIES.R")
print("############################################################")
print("# Convert numbers into string")
print("############################################################")
source("_NUMBERS.R")
print("############################################################")
print("# Generate Document Term Matrix")
print("############################################################")
print("# based on word n-grams")
source("_DTM_WORDS.R")
print("############################################################")
print("# based on character n-grams")
nb_char_grams_max=nb_char_grams_max1
source("_DTM_CHARS.R")
nb_char_grams_max=nb_char_grams_max2
source("_DTM_CHARS.R")
gc()
print("############################################################")
print("# Feature selection")
print("############################################################")
seed=2013 ; Store(seed)
for (DTM_name in c(paste("_DTM_Words_",nb_word_grams_max,"grams",sep=""))) {
print(DTM_name)
seed=2013+seed ; Store(seed)
load(paste("Working_files/",Version,"_",item,DTM_name,".RData",sep=""))
x=data.frame(bin_mat)
source("_FEATURE_SEL_wGLMNET.R")
print("############################################################")
}
for (DTM_name in c(paste("_DTM_Chars_",nb_char_grams_max1,"grams",sep=""),
paste("_DTM_Chars_",nb_char_grams_max2,"grams",sep=""))) {
print(DTM_name)
seed=2013+seed ; Store(seed)
load(paste("Working_files/",Version,"_",item,DTM_name,".RData",sep=""))
x=data.frame(bin_mat)
source("_FEATURE_SEL_wRF.R")
print("############################################################")
}
gc()
print("############################################################")
print("# Principal Component Analysis")
print("############################################################")
for (DTM_name in c(paste("_Small_DTM_Chars_",nb_char_grams_max1,"grams",sep=""),
paste("_Small_DTM_Chars_",nb_char_grams_max2,"grams",sep=""))) {
print(DTM_name)
load(paste("Working_files/",Version,"_",item,DTM_name,".RData",sep=""))
source("_PCA.R")
}
gc()
print("############################################################")
print("# Train Trees")
print("############################################################")
seed=675 ; Store(seed)
DTM_name =paste("_Small2_DTM_Words_",nb_word_grams_max,"grams",sep="")
print(DTM_name)
load(paste("Working_files/",Version,"_",item,DTM_name,".RData",sep=""))
GBM_interaction.depth=8
source("_TRAIN_TREES.R")
gc()
print("############################################################")
print('ADD PROXIES')
print("############################################################")
DTM_name =paste("_Small2_DTM_Words_",nb_word_grams_max,"grams",sep="")
load(paste("Working_files/",Version,"_",item,DTM_name,".RData",sep=""))
load(paste("Working_files/",Version,"_",item,"_PROXIES.RData",sep=""))
for (i in 1:length(x)) x[[i]]<-data.frame(x[[i]],PROXIES[,-1])
DTM_name<-paste(DTM_name,"_PROXIES",sep="")
print(DTM_name)
GBM_interaction.depth=12
source("_TRAIN_TREES.R")
gc()
print("############################################################")
print("# Train Glmnet and SVMs")
print("############################################################")
seed=675 ; Store(seed)
for (DTM_name in c(paste("_Small2_DTM_Words_",nb_word_grams_max,"grams",sep=""),
paste("_Small_DTM_Chars_",nb_char_grams_max1,"grams",sep=""),
paste("_PCA_Small_DTM_Chars_",nb_char_grams_max1,"grams",sep=""),
paste("_Small_DTM_Chars_",nb_char_grams_max2,"grams",sep=""),
paste("_PCA_Small_DTM_Chars_",nb_char_grams_max2,"grams",sep=""))) {
print(DTM_name)
load(paste("Working_files/",Version,"_",item,DTM_name,".RData",sep=""))
source("_TRAIN_NONTREES.R")
}
gc()
print("############################################################")
print("# Group Individual Models")
print("############################################################")
Mod2Blend=data.frame(y=y)
for (DTM_name in c(paste("DTM_Words_",nb_word_grams_max,"grams",sep=""))) {
print(DTM_name)
load(paste("Working_files/",Version,"_",item,"_GLMNET_4fs_",DTM_name,".RData",sep=""))
Mod2Blend<-data.frame(Mod2Blend,GLMNET_4fs$yhatV)
names(Mod2Blend)[ncol(Mod2Blend)]<-paste(DTM_name,c("GLMNET_4fs"),sep="_")
}
for (DTM_name in c(paste("DTM_Chars_",nb_char_grams_max1,"grams",sep=""),
paste("DTM_Chars_",nb_char_grams_max2,"grams",sep=""))) {
print(DTM_name)
load(paste("Working_files/",Version,"_",item,"_RF_4fs_",DTM_name,".RData",sep=""))
Mod2Blend<-data.frame(Mod2Blend,RF_4fs$yhatV)
names(Mod2Blend)[ncol(Mod2Blend)]<-paste(DTM_name,c("RF_4fs"),sep="_")
}
for (DTM_name in c(paste("Small2_DTM_Words_",nb_word_grams_max,"grams",sep=""),
paste("Small2_DTM_Words_",nb_word_grams_max,"grams_PROXIES",sep=""))) {
print(DTM_name)
load(paste("Working_files/",Version,"_",item,"_Trees_w_",DTM_name,".RData",sep=""))
Mod2Blend<-data.frame(Mod2Blend,RF$yhatV,GBM$yhatV)
names(Mod2Blend)[ncol(Mod2Blend)+(-1:0)]<-paste(DTM_name,c("RF","GBM"),sep="_")
}
for (DTM_name in c(paste("Small2_DTM_Words_",nb_word_grams_max,"grams",sep=""),
paste("Small_DTM_Chars_",nb_char_grams_max1,"grams",sep=""),
paste("PCA_Small_DTM_Chars_",nb_char_grams_max1,"grams",sep=""),
paste("Small_DTM_Chars_",nb_char_grams_max2,"grams",sep=""),
paste("PCA_Small_DTM_Chars_",nb_char_grams_max2,"grams",sep=""))) {
print(DTM_name)
load(paste("Working_files/",Version,"_",item,"_NonTrees_w_",DTM_name,".RData",sep=""))
Mod2Blend<-cbind(Mod2Blend,
GLMNET$yhatV,
SVM_lin$yhatV,
SVM_rad$yhatV)
names(Mod2Blend)[ncol(Mod2Blend)+(-2:0)]<-
paste(DTM_name,c("GLMNET","SVM_lin","SVM_rad"),sep="_")
}
save(Mod2Blend,file=paste("Working_files/",Version,"_",item,"_Mod2Blend.RData",sep=""))
print("############################################################")
}
gc()
print("############################################################")
print("# Blend")
#####################################################################
source("CV_NNLS.R")
source("_KFolds.R")
for (item in item_list) {
load(paste("Working_files/",Version,"_",item,"_Mod2Blend.RData",sep=""))
Kb<-Kfolds(896,nb_folds_training,nrow(Mod2Blend))
NNLS<-CV_NNLS(x=Mod2Blend[,-1],y=Mod2Blend$y,
K=Kb,nb_cores=nb_cores,seed=9867)
print(item)
save(NNLS,Kb,file=paste("Working_files/",Version,"_",item,"_NNLS.RData",sep=""))
}
gc()
print("############################################################")
print("# Adjust to optimize Kappa scores")
#####################################################################
source("CV_ADJ.R")
source("_KFolds.R")
for (item in item_list) {
load(paste("Working_files/",Version,"_",item,"_Mod2Blend.RData",sep=""))
load(paste("Working_files/",Version,"_",item,"_NNLS.RData",sep=""))
Kc<-Kfolds(896,nb_folds_training,length(NNLS$yhatV))
ADJ1<-CV_ADJ(x=NNLS$yhatV,y=Mod2Blend$y,K=Kc,d=1,nb_cores=nb_cores,seed=seed)
save(ADJ1,Kc,file=paste("Working_files/",Version,"_",item,"_ADJ1.RData",sep=""))
print(item)
}
for (item in item_list) {
load(paste("Working_files/",Version,"_",item,"_Mod2Blend.RData",sep=""))
load(paste("Working_files/",Version,"_",item,"_NNLS.RData",sep=""))
Kc<-Kfolds(896,nb_folds_training,length(NNLS$yhatV))
ADJ2<-CV_ADJ(x=NNLS$yhatV,y=Mod2Blend$y,K=Kc,d=2,nb_cores=nb_cores,seed=seed)
save(ADJ2,Kc,file=paste("Working_files/",Version,"_",item,"_ADJ2.RData",sep=""))
print(item)
}