-
Notifications
You must be signed in to change notification settings - Fork 2
/
analysis_usd.R
228 lines (152 loc) · 9.63 KB
/
analysis_usd.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
### Filter the best_usd result_usd
library("dplyr")
# HOMO model
result_usd <- as.tbl(result_HOMO_USD)
min_value <- min(result_usd$RMSE)
best_usd <- filter(result_usd, RMSE == min(result_usd$RMSE))
for (i in 1:length(result_usd$RMSE)) {
if(result_usd$RMSE[i]== min_value){
row_select <- i
}
}
predictor_order <- best_usd$Predictor_Order
neurons <- best_usd$Neurons
activation_func <- best_usd$Activation_Function
learning_rate <- best_usd$Learning_Rate
## Getting Error and actual value from the result_usd LIST to plot
predict_value<- Result_USD_HOMO_LIST[[row_select]][[5]][[1]] ##[[the row number in result_HOMO_*]][[5]][[1]]
predict_error <- Result_USD_HOMO_LIST[[row_select]][[5]][[2]] ##[[the row number in result_HOMO_*]][[5]][[2]]
homo_model1_usd <- Result_USD_HOMO_LIST[[row_select]][[5]][[6]][[1]] ##[[the row number in result_HOMO_*][[5]][[6]][[1]]
homo_model2_usd <- Result_USD_HOMO_LIST[[row_select]][[5]][[6]][[2]] ##[[the row number in result_HOMO_*][[5]][[6]][[2]]
homo_model3_usd <- Result_USD_HOMO_LIST[[row_select]][[5]][[6]][[3]] ##[[the row number in result_HOMO_*][[5]][[6]][[3]]
#******* Training 70% ******
predict_value<- Result_USD_HOMO_LIST_70[[row_select]][[5]][[1]] ##[[the row number in result_HOMO_*]][[5]][[1]]
predict_error <- Result_USD_HOMO_LIST_70[[row_select]][[5]][[2]] ##[[the row number in result_HOMO_*]][[5]][[2]]
homo_model1_usd <- Result_USD_HOMO_LIST_70[[row_select]][[5]][[6]][[1]] ##[[the row number in result_HOMO_*][[5]][[6]][[1]]
homo_model2_usd <- Result_USD_HOMO_LIST_70[[row_select]][[5]][[6]][[2]] ##[[the row number in result_HOMO_*][[5]][[6]][[2]]
homo_model3_usd <- Result_USD_HOMO_LIST_70[[row_select]][[5]][[6]][[3]] ##[[the row number in result_HOMO_*][[5]][[6]][[3]]
#******* Training 80% ******
predict_value<- Result_USD_HOMO_LIST_80[[row_select]][[5]][[1]] ##[[the row number in result_HOMO_*]][[5]][[1]]
predict_error <- Result_USD_HOMO_LIST_80[[row_select]][[5]][[2]] ##[[the row number in result_HOMO_*]][[5]][[2]]
homo_model1_usd <- Result_USD_HOMO_LIST_80[[row_select]][[5]][[6]][[1]] ##[[the row number in result_HOMO_*][[5]][[6]][[1]]
homo_model2_usd <- Result_USD_HOMO_LIST_80[[row_select]][[5]][[6]][[2]] ##[[the row number in result_HOMO_*][[5]][[6]][[2]]
homo_model3_usd <- Result_USD_HOMO_LIST_80[[row_select]][[5]][[6]][[3]] ##[[the row number in result_HOMO_*][[5]][[6]][[3]]
test_data <- data_set[[1]][[predictor_order-2]][[3]] ## [[currency]][[predictor_order-2]][[1:6]] 1.Traindata,
validate_data <- data_set[[1]][[predictor_order-2]][[2]] # 2.Validate Data 3. TestData 4.validate_date 5. test_Date
# 6. actual whole dataset
#Applying three models to valadatation dataset
non_normalize <- data_set[[1]][[predictor_order - 2]][[6]]
model_results <- neuralnet::compute(homo_model1_usd, test_data[1:predictor_order])
predicted_oneDayhead <- denormalized(model_results$net.result,non_normalize)
model_results2 <- neuralnet::compute(homo_model2_usd, test_data[1:predictor_order])
predicted_oneDayhead2 <- denormalized(model_results2$net.result,non_normalize)
model_results3 <- neuralnet::compute(homo_model3_usd, test_data[1:predictor_order])
predicted_oneDayhead3 <- denormalized(model_results3$net.result,non_normalize)
all_predicted <- cbind(predicted_oneDayhead,predicted_oneDayhead2,predicted_oneDayhead3)
all_predicted <-as.data.frame(all_predicted)
actual <- denormalized(test_data[,predictor_order+1],non_normalize)
min_value <-apply(all_predicted,1, min)
max_value <- apply(all_predicted,1,max)
mean_value <- apply(all_predicted,1,mean)
error_min <- actual - min_value
error_max <- actual - max_value
error_mean <- actual - mean_value
error_all_after_fusion <- as.data.frame(cbind(error_min,error_max,error_mean))
names(error_all_after_fusion) <- c("MIN","MAX","MEAN")
rmse_min <- rmse(error_all_after_fusion$MIN)
mae_min <- mae(error_all_after_fusion$MIN)
rmse_max <- rmse(error_all_after_fusion$MAX)
mae_max <- mae(error_all_after_fusion$MAX)
rmse_mean <- rmse(error_all_after_fusion$MEAN)
mae_mean<- mae(error_all_after_fusion$MEAN)
rmse_rate <- min(rmse_min,rmse_max,rmse_mean)
if(rmse_rate == rmse_max ){
final_result_test_homo = list(unname(max_value),unname(error_max),"MAX",rmse_rate,mae_max)
}
if(rmse_rate == rmse_min ){
final_result_test_homo = list(unname(min_value),unname(error_min),"MIN",rmse_rate,mae_min)
}
if(rmse_rate == rmse_mean){
final_result_test_homo = list(unname(mean_value),unname(error_mean),"MEAN",rmse_rate,mae_mean)
}
######################################## Hetro Model
result_hetro <- as.tbl(result_HETRO_USD)
min_hetro_value <- min(result_hetro$RMSE)
best_hetro_usd <- filter(result_hetro, RMSE == min_hetro_value)
for (i in 1:length(result_hetro$RMSE)) {
if(result_hetro$RMSE[i]== min_hetro_value){
row_select_hetro <- i
}
}
predictor_order_hetro <- best_hetro_usd$Predictor.Order
neurons_hetro <- best_hetro_usd$Neurons
activation_func_htero <- best_hetro_usd$Activation.Func
learning_rate_hetro <- best_hetro_usd$Learning.Rate
## Getting Error and actual value from the result_usd LIST to plot
predict_value<- Result_USD_HETRO_LIST[[row_select]][[5]][[1]] ##[[the row number in result_HOMO_*]][[5]][[1]]
predict_error <- Result_USD_HETRO_LIST[[row_select]][[5]][[2]] ##[[the row number in result_HOMO_*]][[5]][[2]]
hetro_model1_usd <- Result_USD_HETRO_LIST[[row_select]][[5]][[6]][[1]] ##[[the row number in result_HOMO_*][[5]][[6]][[1]]
hetro_model2_usd <- Result_USD_HETRO_LIST[[row_select]][[5]][[6]][[2]] ##[[the row number in result_HOMO_*][[5]][[6]][[2]]
hetro_model3_usd <- Result_USD_HETRO_LIST[[row_select]][[5]][[6]][[3]] ##[[the row number in result_HOMO_*][[5]][[6]][[3]]
#******* Training 70% ******
predict_value<- Result_USD_HETRO_LIST_70[[row_select]][[5]][[1]] ##[[the row number in result_HOMO_*]][[5]][[1]]
predict_error <- Result_USD_HETRO_LIST_70[[row_select]][[5]][[2]] ##[[the row number in result_HOMO_*]][[5]][[2]]
hetro_model1_usd <- Result_USD_HETRO_LIST_70[[row_select]][[5]][[6]][[1]] ##[[the row number in result_HOMO_*][[5]][[6]][[1]]
hetro_model2_usd <- Result_USD_HETRO_LIST_70[[row_select]][[5]][[6]][[2]] ##[[the row number in result_HOMO_*][[5]][[6]][[2]]
hetro_model3_usd <- Result_USD_HETRO_LIST_70[[row_select]][[5]][[6]][[3]] ##[[the row number in result_HOMO_*][[5]][[6]][[3]]
#******* Training 80% ******
predict_value<- Result_USD_HETRO_LIST_80[[row_select]][[5]][[1]] ##[[the row number in result_HOMO_*]][[5]][[1]]
predict_error <- Result_USD_HETRO_LIST_80[[row_select]][[5]][[2]] ##[[the row number in result_HOMO_*]][[5]][[2]]
hetro_model1_usd <- Result_USD_HETRO_LIST_80[[row_select]][[5]][[6]][[1]] ##[[the row number in result_HOMO_*][[5]][[6]][[1]]
hetro_model2_usd <- Result_USD_HETRO_LIST_80[[row_select]][[5]][[6]][[2]] ##[[the row number in result_HOMO_*][[5]][[6]][[2]]
hetro_model3_usd <- Result_USD_HETRO_LIST_80[[row_select]][[5]][[6]][[3]] ##[[the row number in result_HOMO_*][[5]][[6]][[3]]
test_data <- data_set[[1]][[predictor_order-2]][[3]] ## [[currency]][[predictor_order-2]][[1:6]] 1.Traindata,
validate_data <- data_set[[1]][[predictor_order-2]][[2]] # 2.Validate Data 3. TestData 4.validate_date 5. Test_Date
# 6. actual whole dataset
#Applying three models to valadatation dataset
non_normalize <- data_set[[1]][[predictor_order - 2]][[6]]
# Model 1
model_results <- neuralnet::compute(hetro_model1_usd, test_data[1:predictor_order])
predicted_oneDayhead <- denormalized(model_results$net.result,non_normalize)
# Model 2
test_input_array_1 <- list()
for (i in 1:predictor_order){
input_day <- as.matrix(test_data[,i])
test_input_array_1 <- c(test_input_array_1,input_day)
}
test_input_array_1 <- as.numeric(test_input_array_1)
test_input <- array( test_input_array_1, dim=c(dim(input_day),predictor_order) )
model_results2 <- predictr(hetro_model2_usd, test_input)
predicted_oneDayhead2 <- denormalized(model_results2,non_normalize)
## Model 3
test_input <- as.matrix(test_data[,1:predictor_order])
model_results3 <- predict(hetro_model3_usd,test_input)
model_results3 <- as.vector(model_results3)
predicted_oneDayhead3 <- denormalized(model_results3,non_normalize)
all_predicted <- cbind(predicted_oneDayhead,predicted_oneDayhead2,predicted_oneDayhead3)
all_predicted <-as.data.frame(all_predicted)
actual <- denormalized(test_data[,predictor_order+1],non_normalize)
min_value <-apply(all_predicted,1, min)
max_value <- apply(all_predicted,1,max)
mean_value <- apply(all_predicted,1,mean)
error_min <- actual - min_value
error_max <- actual - max_value
error_mean <- actual - mean_value
error_all_after_fusion <- as.data.frame(cbind(error_min,error_max,error_mean))
names(error_all_after_fusion) <- c("MIN","MAX","MEAN")
rmse_min <- rmse(error_all_after_fusion$MIN)
mae_min <- mae(error_all_after_fusion$MIN)
rmse_max <- rmse(error_all_after_fusion$MAX)
mae_max <- mae(error_all_after_fusion$MAX)
rmse_mean <- rmse(error_all_after_fusion$MEAN)
mae_mean<- mae(error_all_after_fusion$MEAN)
rmse_rate <- min(rmse_min,rmse_max,rmse_mean)
if(rmse_rate == rmse_max ){
final_result_test_hetro = list(unname(max_value),unname(error_max),"MAX",rmse_rate,mae_max)
}
if(rmse_rate == rmse_min ){
final_result_test_hetro = list(unname(min_value),unname(error_min),"MIN",rmse_rate,mae_min)
}
if(rmse_rate == rmse_mean){
final_result_test_hetro = list(unname(mean_value),unname(error_mean),"MEAN",rmse_rate,mae_mean)
}