-
Notifications
You must be signed in to change notification settings - Fork 0
/
regression_classification.R
230 lines (165 loc) · 11.6 KB
/
regression_classification.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
source("Seed.R")
## === START: MCAR Data preparartion ===
# Copying the working set
mcar_regcla_set = mcar_working_set
# reduce size of dataset for classification and persue classification for cities based on distances:
mcar_regcla_set = mcar_regcla_set[, c("X003_Postcode","X004_City","X010_Lon", "X011_Lat", "X012_NoRooms", "X099_DistanceCathedralLeSeu", "X100_DistanceAlcudiaOldTown", "X098_DistanceSerraDeTramuntana")]
# Now we're doing classification w/ the imputed values. So we include distToBeach_imp
mcar_regcla_set = cbind(mcar_regcla_set,distToBeach_imp)
# Having Postcode as numeric proved as a more performant solution than having a factor w/ a huge amount of levels
mcar_regcla_set[ , "X003_Postcode"] <- as.numeric(mcar_regcla_set[ , "X003_Postcode"])
mcar_regcla_set[ , "X012_NoRooms"] <- as.numeric(mcar_regcla_set[ , "X012_NoRooms"])
#Updating the levels. Otherwise error: Can't have empty classes in y
mcar_regcla_set[ , "X004_City"] <- as.character(mcar_regcla_set[ , "X004_City"])
mcar_regcla_set[ , "X004_City"] <- as.factor(mcar_regcla_set[ , "X004_City"])
#Copying the original dataset (for means of error calculation later on)
regcla_working_set <- working_set
regcla_working_set[ , "X004_City"] <- as.character(regcla_working_set[ , "X004_City"])
#Drop NA rows for model creation
mcar_regcla_set_woNA <- na.omit(mcar_regcla_set)
# Creating an index to identify the missing values
mcar_regcla_set_names <- colnames(mcar_regcla_set)
IndexOfMissing_city <- which(is.na(mcar_regcla_set$X004_City))
str(mcar_regcla_set_woNA)
str(mcar_regcla_set)
## === END: MCAR Data preparartion ===
##=== START: Classification via Tree for MCAR===
tr_regcla_MCAR = rpart(X004_City ~ ., data=mcar_regcla_set_woNA)
#####Predictions
mcar_tree_pred_all = predict(tr_regcla_MCAR, mcar_regcla_set[,!(mcar_regcla_set_names %in% "X004_City")], type = "class")
mcar_tree_pred_all_error <- mean((as.data.frame(mcar_tree_pred_all)[,1]) != regcla_working_set$X004_City) #NA, in case of missing values
mcar_tree_pred_ten = predict(tr_regcla_MCAR, mcar_regcla_set[IndexOfMissing_city,!(mcar_regcla_set_names %in% "X004_City")], type = "class")
mcar_tree_pred_ten_error <- mean((as.data.frame(mcar_tree_pred_ten)[,1]) != regcla_working_set$X004_City[IndexOfMissing_city]) #NA, in case of missing values
#== Some further checks:
cbind(as.character(as.data.frame(mcar_tree_pred_ten)[,1]),as.character(regcla_working_set$X004_City[IndexOfMissing_city]))
#==
#== Plotting:
setEPS()
postscript("Cla_City_Imputed_mcar_Tree.eps")
plot(tr_regcla_MCAR, main="MCAR Classification Tree for City", sub=paste("Prediction Error for all: ",round(mcar_tree_pred_all_error,4)," Prediction error for ten: ",round(mcar_tree_pred_ten_error,4)))
text(tr_regcla_MCAR, pretty = 0, cex=0.8)
dev.off()
#==
##=== END: Classification via Tree ===
# Interesting: Setting the factorial feature "postcode" as numerical brings an advantage of 0.1
##=== START: Classification via Forest for MCAR===
#####Model creation
mcar_forest_regcla = randomForest(X004_City ~ ., data=mcar_regcla_set_woNA, ntree=50)
#####Predictions
mcar_forest_pred_all = predict(mcar_forest_regcla, mcar_regcla_set[,!(mcar_regcla_set_names %in% "X004_City")], type = "class")
mcar_forest_pred_all_error <- mean((as.data.frame(as.character(mcar_forest_pred_all))[,1]) != regcla_working_set$X004_City) #NA, in case of missing values
mcar_forest_pred_ten = predict(mcar_forest_regcla, mcar_regcla_set[IndexOfMissing_city,!(mcar_regcla_set_names %in% "X004_City")], type = "class")
mcar_forest_pred_ten_error <- mean((as.data.frame(as.character(mcar_forest_pred_ten))[,1]) != regcla_working_set$X004_City[IndexOfMissing_city]) #NA, in case of missing values
#== Some further checks:
cbind(as.character(as.data.frame(as.character(mcar_forest_pred_ten))[,1]),regcla_working_set$X004_City[IndexOfMissing_city])
#==
#== Plotting:
setEPS()
postscript("Cla_City_Imputed_mcar_Forest.eps")
varImpPlot(mcar_forest_regcla, main="Classification via Forest for City", sub=paste("Prediction Error for all: ",round(mcar_forest_pred_all_error,4)," Prediction error for ten: ",round(mcar_forest_pred_ten_error,4))) # Interpretation necessary
dev.off()
#==
##=== END: Classification via Forest ===#
## === START: MAR Data preparartion ===
# Copying the working set
# "X004_City" caused massive performance problems here. Probably it has too many levels.
# So we removed it
mar_regcla_set = mar_working_set[, c("X003_Postcode", "X010_Lon", "X011_Lat", "X102_DistanceToBeach", "X099_DistanceCathedralLeSeu", "X100_DistanceAlcudiaOldTown", "X098_DistanceSerraDeTramuntana")]
# Now we're doing classification w/ the imputed values. So we include rooms_imp
mcar_regcla_set = cbind(mcar_regcla_set,rooms_imp)
#Updating the levels. Otherwise error: Can't have empty classes in y
mar_regcla_set[ , "X003_Postcode"] <- as.character(mar_regcla_set[ , "X003_Postcode"])
mar_regcla_set[ , "X003_Postcode"] <- as.factor(mar_regcla_set[ , "X003_Postcode"])
# Creating an index to identify the missing values
mar_regcla_set_names <- colnames(mar_regcla_set)
IndexOfMissing_postcode <- which(is.na(mar_regcla_set$X003_Postcode))
#Drop NA rows for model creation
mar_regcla_set_woNA <- na.omit(mar_regcla_set)
## === END: MAR Data preparation ===
##=== START: Classification via Tree for MAR===
tr_regcla_MAR = rpart(X003_Postcode ~ ., data=mar_regcla_set_woNA)
#####Predictions
mar_tree_pred_all <- predict(tr_regcla_MAR, mar_regcla_set[,!(mar_regcla_set_names %in% "X003_Postcode")], type="class")
mar_tree_pred_all_error <- mean((as.data.frame(mar_tree_pred_all)[,1]) != as.character(regcla_working_set$X003_Postcode)) #NA, in case of missing values
mar_tree_pred_ten <- predict(tr_regcla_MAR, mar_regcla_set[IndexOfMissing_postcode,!(mar_regcla_set_names %in% "X003_Postcode")], type="class")
mar_tree_pred_ten_error <- mean((as.data.frame(mar_tree_pred_ten)[,1]) != as.character(regcla_working_set$X003_Postcode[IndexOfMissing_postcode])) #NA, in case of missing values
#== Some further checks:
cbind(as.character(as.data.frame(mar_tree_pred_ten)[,1]),as.character(regcla_working_set$X003_Postcode[IndexOfMissing_postcode]))
#==
#== Plotting:
setEPS()
postscript("Cla_Postcode_Imputed_mar_Tree.eps")
plot(tr_regcla_MAR, main="MAR Classification Tree for Postcode", sub=paste("Prediction Error for all: ",round(mar_tree_pred_all_error,4),"Prediction error for ten: ",round(mar_tree_pred_ten_error,4)))
text(tr_regcla_MAR, pretty = 0, cex=0.8)
dev.off()
##=== END: Classification via Tree ===
##=== START: Classification via Forest for MAR===
#####Model Creation
mar_forest_regcla = randomForest(X003_Postcode ~., data=mar_regcla_set_woNA, ntree = 50)
#####Prediction
mar_forest_pred_all = predict(mar_forest_regcla, mar_regcla_set[,!(mar_regcla_set_names %in% "X003_Postcode")], type = "class")
mar_forest_pred_all_error <- mean((as.data.frame(mar_forest_pred_all)[,1]) != as.numeric(regcla_working_set$X003_Postcode))
mar_forest_pred_ten <- predict(mar_forest_regcla, mar_regcla_set[IndexOfMissing_postcode,!(mar_regcla_set_names %in% "X003_Postcode")], type = "class")
mar_forest_pred_ten_error <- mean(as.character(as.data.frame(mar_forest_pred_ten)[,1]) != (as.character(as.data.frame(regcla_working_set$X003_Postcode[IndexOfMissing_postcode])[,1])))
#== Some further checks:
cbind(as.character(as.data.frame(mar_forest_pred_ten)[,1]),(as.character(as.data.frame(regcla_working_set$X003_Postcode[IndexOfMissing_postcode])[,1])))
#==
#== Plotting:
setEPS()
postscript("Cla_Postcode_Imputed_mar_Forest.eps")
varImpPlot(mar_forest_regcla, main="Classification via Forest for Postcode", sub=paste("Prediction Error for all: ",round(mar_forest_pred_all_error,4)," Prediction error for ten: ",round(mar_forest_pred_ten_error,4))) # Interpretation necessary) # Interpretation necessary
dev.off()
#==
##=== END: Classification via Forest ===
## === START: MNAR Data preparation ===
# Copying the working set
# reduce size of dataset for classification and persue classification for PayOptions manly based on distances:
mnar_regcla_set = mnar_working_set[, c("X046_PayOptions", "X012_NoRooms", "X102_DistanceToBeach", "X099_DistanceCathedralLeSeu", "X098_DistanceSerraDeTramuntana")]
# Set Pay Options as factor since we want to use it for classification
mnar_regcla_set[ , "X046_PayOptions"] <- as.factor(mnar_regcla_set[ , "X046_PayOptions"])
# Now we're doing classification w/ the imputed values. So we include distToBeach_imp
mnar_regcla_set = cbind(mnar_regcla_set,old_town_imp)
#Updating the levels. Otherwise error: Can't have empty classes in y
mnar_regcla_set[ , "X046_PayOptions"] <- as.character(mnar_regcla_set[ , "X046_PayOptions"])
mnar_regcla_set[ , "X046_PayOptions"] <- as.factor(mnar_regcla_set[ , "X046_PayOptions"])
# Creating an index to identify the missing values
mnar_regcla_set_names <- colnames(mnar_regcla_set)
IndexOfMissing_PayOptions <- which(is.na(mnar_regcla_set$X046_PayOptions))
# Drop NA rows for model creation
mnar_regcla_set_woNA <- na.omit(mnar_regcla_set)
## === END: MNAR Data preparartion ===
##=== START: Classification via Tree for MNAR===
#####Model Creation
tr_regcla_MNAR = rpart(X046_PayOptions ~ ., data=mnar_regcla_set_woNA)
#####Prediction
mnar_tree_pred_all = predict(tr_regcla_MNAR, mnar_regcla_set[,!(mnar_regcla_set_names %in% "X046_PayOptions")], type = "class")
mnar_tree_pred_all_error <- mean((as.data.frame(mnar_tree_pred_all)[,1]) != as.character(regcla_working_set$X046_PayOptions))
mnar_tree_pred_ten = predict(tr_regcla_MNAR, mnar_regcla_set[IndexOfMissing_PayOptions,!(mnar_regcla_set_names %in% "X046_PayOptions")], type = "class")
mnar_tree_pred_ten_error <- mean(as.character(as.data.frame(mnar_tree_pred_ten)[,1]) != as.character(regcla_working_set$X046_PayOptions[IndexOfMissing_PayOptions]))
#== Some further checks:
cbind(as.character(as.data.frame(mnar_tree_pred_ten)[,1]),as.character(as.data.frame(regcla_working_set$X046_PayOptions[IndexOfMissing_PayOptions])[,1]))
#==
#== Plotting:
setEPS()
postscript("Cla_PayOptions_Imputed_mnar_Tree.eps")
plot(tr_regcla_MNAR, main="MNAR Classification Tree for Pay Options", sub=paste("Prediction Error for all: ",round(mnar_tree_pred_all_error,4),"Prediction error for ten: ",round(mnar_tree_pred_ten_error,4)))
text(tr_regcla_MNAR, pretty = 0, cex=0.8)
dev.off()
##=== END: Classification via Tree ===
##=== START: Classification via Forest for MNAR===
#####Model Creation
mnar_forest_regcla = randomForest(X046_PayOptions ~ ., data=mnar_regcla_set_woNA, ntree = 50, na.action=na.omit)
#####Prediction
mnar_forest_pred_all = predict(mnar_forest_regcla, mnar_regcla_set[,!(mnar_regcla_set_names %in% "X046_PayOptions")], type = "class")
mnar_forest_pred_all_error <- mean(as.character(as.data.frame(mnar_forest_pred_all)[,1]) != as.character(as.data.frame(regcla_working_set$X046_PayOptions)[,1]))
mnar_forest_pred_ten = predict(mnar_forest_regcla, mnar_regcla_set[IndexOfMissing_PayOptions,!(mnar_regcla_set_names %in% "X046_PayOptions")], type = "class")
mnar_forest_pred_ten_error <- mean((as.data.frame(mnar_forest_pred_ten)[,1]) != as.character(as.data.frame(regcla_working_set$X046_PayOptions[IndexOfMissing_PayOptions])[,1]))
#== Some further checks:
cbind(as.character((as.data.frame(mnar_forest_pred_ten)[,1])),as.character(as.data.frame(regcla_working_set$X046_PayOptions[IndexOfMissing_PayOptions])[,1]))
#==
#== Plotting:
setEPS()
postscript("Cla_PayOptions_Imputed_mnar_Tree.eps")
varImpPlot(mnar_forest_regcla, main="Classification via Forest for Pay Options", sub=paste("Prediction Error for all: ",round(mnar_forest_pred_all_error,4)," Prediction error for ten: ",round(mnar_forest_pred_ten_error,4)))
dev.off()
##=== END: Classification via Forest ===