-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathSSFCM_plots.R
228 lines (195 loc) · 8.28 KB
/
SSFCM_plots.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
#ADICIONAR:
# KFCM
#SSFCM_MACARIO
#WSSFCM_KONG
#SSFCM_LI
#SSKFCM
setwd("C:/Users/Gerson/Documents")
library(ggplot2)
library(caret)
library(e1071)
library(gridExtra)
source("SSFCM_Pedrycz.R")
source("SSFCM_Pedrycz_Fuzzy.R")
source("FCM_function.R")
#Lendo DB
#db = read.csv("AlgoritmosAntigos/DataSets/dadosGerados_FCM2_SS.csv")
#db = db[,2:4]
#db2 = read.csv("AlgoritmosAntigos/DataSets/dadosGerados_FCM2.csv")
#db = read.csv("AlgoritmosAntigos/DataSets/dadosGeradosFCM2_SS_90.csv")
#db = db[,2:4]
#db2 = read.csv("AlgoritmosAntigos/DataSets/dadosGerados_FCM2.csv")
db2 = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/BG_10k/BarsGaussAN0_10000.csv")
db2 = db2[4000:5000,]
db = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/BG_10k/BarsGaussAN0_1000_SS.csv")
db = db[,2:4]
plot(db2[,1:2])
#db = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK3D2_10k/MixGaussiansK3D2N0_10000_SS.csv")
#db = db[,2:4]
#db2 = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK3D2_10k/MixGaussiansK3D2N0_10000.csv")
#db2 = db2[1:1000,]
#db = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK3D2_10k/MixGaussiansK3D2N0_10000_SS.csv")
#db = db[,2:4]
#db2 = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK3D2_10k/MixGaussiansK3D2N0_10000.csv")
#db2 = db2[1:1000,]
#db = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK3D3_10k/MixGaussiansK3D3N0_1000_SS.csv")
#db = db[,2:4]
#db2 = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK3D3_10k/MixGaussiansK3D3N0_10000.csv")
#db2 = db2[1:1000,]
#db = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK4D2_10k/MixGaussiansK4D2N0_1000_SS.csv")
#db = db[,2:4]
#db2 = read.csv("OnDemand-master/DS_Datasets/Synthetic/Stationary/MGK4D2_10k/MixGaussiansK4D2N0_10000.csv")
#db2 = db2[1:1000,]
#Quantidade de grupos (Na verdade k deve ser igual a quantidade de classes)
dim = ncol(db)-1
N = nrow(db)
#Renomeando as colunas
colnames(db) = c(1:dim,"class")
#Encontrando a qtd de classes (ARRUMAR MELHOR DEPOIS)
b = db$class
aux = b > 0
b2 = b
b2[aux] = 1
#Encontrando a quantidade de classes
rotulados = as.factor(db$class[aux])
classes = as.numeric(names(table(rotulados)))
nclasses = length(classes)
#threshold
th = 0.001
#Valor alpha do algoritmo
alpha = 0.99
########################CRIANDO PALETAS DE CORES######################################
CORES = NULL
CORES[[2]] = c("gray","red")
CORES[[3]] = c("gray","red","blue")
CORES[[4]] = c("gray","red","blue","green")
CORES[[5]] = c("gray","red","blue","green","yellow")
CORES[[1]] = "gray"
CORES2 = NULL
CORES2[[2]] = c("red","blue")
CORES2[[3]] = c("red","blue","green")
CORES2[[4]] = c("red","blue","green","yellow")
CORES2[[1]] = "red"
CORES3 = NULL
CORES3[[2]] = c("red","black")
CORES3[[3]] = c("red","blue","black")
CORES3[[4]] = c("red","blue","green","black")
CORES3[[5]] = c("red","blue","green","yellow","black")
CORES3[[1]] = "black"
##############################ABRINDO O PDF###########################################
#pdf("ComparacaoResultados_10_1_90.PDF")
##############################Apenas plots para demonstrar############################
family = as.factor(db[,dim+1])
tam = length(table(family))
palette(CORES[[tam]])
#plot(db[,1:2], pch = c(3, 15:(15+tam-2))[family], col=family, xlim=c(15,125), ylim = c(15,125))
plot(db[,1:2], pch = c(3, 15:(15+tam-2))[family], col=family, main = "Distribuição dos dados")
#Plotando o supervisionado
family = as.factor(db2[,dim+1])
tam = length(table(family))
palette(CORES[[tam]])
#plot(db[,1:2], pch = c(3, 15:(15+tam-2))[family], col=family, xlim=c(15,125), ylim = c(15,125))
plot(db2[,1:2], pch = c(3, 15:(15+tam-2))[family], col=family, main = "Distribuição dos dados - Supervisionado")
dbc = db[db$class >= 1,]
family = as.factor(dbc[,dim+1])
tam = length(table(family))
palette(CORES2[[tam]])
plot(dbc[,1:2], pch = c(15:(15+tam-1))[family], col=family, main = "Distribuição dos dados rotulados")
#plot(db2[,1:2], pch = c(15, 16, 17)[family], col=family, xlim=c(15,125), ylim = c(15,125))
###########################################################################################
################OBTENDO CLUSTERIZAÇÕES##################
############teste
resultado = read.csv("BarsGaussAN0_200_MiC.csv")
resultado = resultado[,2:7]
V = matrix(ncol = 2, nrow = 4)
V = teste2$centers - 10
#V[1,] = c(-4,2)
#V[2,] = c(1,4)
#V[3,] = c(0,4)
#V[4,] = c(4,-2)
ssfcm_mahalanobis_fuzzy = ssfcm_pedrycz_fuzzy(resultado[,1:2],V,resultado[,3:ncol(resultado)],alpha,"mahalanobis",th,200)
ssfcm_euclidian_fuzzy = ssfcm_pedrycz_fuzzy(resultado[,1:2],V,resultado[,3:ncol(resultado)],alpha,"euclidian",th,200)
ssfcm_mahalanobis = ssfcm_euclidian_fuzzy
ssfcm_mahalanobis = ssfcm_mahalanobis_fuzzy
#Plotando resultado
classificacao = vector(length = nrow(resultado))
res = resultado[,3:ncol(resultado)]
for(i in 1:nrow(res)){
for(j in 1:3){
if(res[i,j] > 0){
res[i,j] = j
}
else{
res[i,j] = 0
}
}
}
for(i in 1:200){
classificacao[i] = sum(res[i,])
}
family = as.factor(classificacao)
tam = length(table(family))
palette(CORES[[tam]])
#plot(db[,1:2], pch = c(3, 15:(15+tam-2))[family], col=family, xlim=c(15,125), ylim = c(15,125))
plot(resultado[,1:2], pch = c(3, 15:(15+tam-2))[family], col=family, main = "Distribuição dos dados")
db = resultado[,1:2]
db[,3] = ssfcm_euclidian_fuzzy$cluster
family = as.factor(db[,3])
tam = length(table(family))
palette(CORES2[[tam]])
plot(db[,1:2], pch = c(15:(15+tam-1))[family], col=family, main = a[[i]]$call )
###################
ssfcm_mahalanobis = ssfcm_pedrycz(db,NULL,NULL,alpha,"mahalanobis",th,200)
ssfcm_euclidian = ssfcm_pedrycz(db,NULL,NULL,alpha,"euclidian",0.001,200)
fcm_euclidian = cmeans(db[,1:dim], inicializeV(dim,nclasses,db), iter.max=100, verbose=FALSE, dist="euclidean", method="cmeans", m=2, rate.par = NULL)
fcm_mahalanobis = fcm_function(db[,1:dim],inicializeV(dim,nclasses,db),"mahalanobis",0.001,200)
#fcm_euclidian2 = fcm_function(db[,1:dim],inicializeV(dim,nclasses,db),"euclidian",0.001,200)
##############FAZENDO PLOTS DOS RESULTADOS###########################################
qtdAlgs = 4
a = NULL
ssfcm_mahalanobis$call = "SSFCM Mahalanobis"
a[[1]] = ssfcm_mahalanobis
ssfcm_euclidian$call = "SSFCM Euclidian"
a[[2]] = ssfcm_euclidian
fcm_mahalanobis$call = "FCM Mahalanobis"
a[[3]] = fcm_mahalanobis
fcm_euclidian$call = "FCM Euclidian"
a[[4]] = fcm_euclidian
#ssfcm_mahalanobis$call = "FCM Euclidian2"
#a[[5]] = fcm_euclidian2
for(i in 1:qtdAlgs){
#Plot apenas com os valores classificados
family = as.factor(a[[i]]$cluster)
tam = length(table(family))
palette(CORES2[[tam]])
plot(db[,1:2], pch = c(15:(15+tam-1))[family], col=family, main = a[[i]]$call )
#Plot dos resultados com os valores dos centroides
V = cbind(a[[i]]$centers,length(a[[i]]$size)+1)
colnames(V) = colnames(db)
classificados = cbind(db[,1:dim],a[[i]]$cluster)
colnames(classificados) = colnames(V)
classificados = rbind(classificados,V)
family = as.factor(classificados[,dim+1])
tam = length(table(family))
palette(CORES3[[tam]])
plot(classificados[,1:2], pch = c(15:(15+tam-2), 3)[family], col=family, main = a[[i]]$call)
}
###################Fazendo matrix de confusão para avaliar resultados#####################
mcFCM_euclidian = confusionMatrix(as.factor(fcm_euclidian$cluster), as.factor(db2[,dim+1]))
#mcFCM_euclidian2 = confusionMatrix(as.factor(fcm_euclidian2$cluster), as.factor(db2[,dim+1]))
mcFCM_mahalanobis = confusionMatrix(as.factor(fcm_mahalanobis$cluster), as.factor(db2[,dim+1]))
mcSSFCM_euclidian = confusionMatrix(as.factor(ssfcm_euclidian$cluster), as.factor(db2[,dim+1]))
mcSSFCM_mahalanobis = confusionMatrix(as.factor(ssfcm_mahalanobis$cluster), as.factor(db2[,dim+1]))
##########################Adicionando os resultados a uma estrutura######################
resultados = cbind("SSFCM Mahalanobis",mcSSFCM_mahalanobis$overall["Accuracy"])
resultados = rbind(resultados,cbind("SSFCM Euclidian",mcSSFCM_euclidian$overall["Accuracy"]))
resultados = rbind(resultados,cbind("FCM Mahalanobis",mcFCM_mahalanobis$overall["Accuracy"]))
resultados = rbind(resultados,cbind("FCM Euclidian",mcFCM_euclidian$overall["Accuracy"]))
#resultados = rbind(resultados,cbind("FCM2 Euclidian",mcFCM_euclidian2$overall["Accuracy"]))
#Renomeando linhas e colunas
row.names(resultados) = c(1:qtdAlgs)
colnames(resultados) = c("Algoritmo","Accuracy")
#plot.new()
frame()
grid.table(resultados)
dev.off()