Blame view
bnlearn.R
7.24 KB
8375af764 add all code 11-7... |
1 2 |
setwd("/home/qja0428/Dropbox/research/Yoo") |
94c6dd68c 11-29-2016 |
3 |
#setwd("/Users/jinganqu/Dropbox/research/Yoo/code") |
8375af764 add all code 11-7... |
4 5 6 7 |
library(bnlearn) data(learning.test) |
8375af764 add all code 11-7... |
8 |
|
94c6dd68c 11-29-2016 |
9 |
dataset = learning.test |
8375af764 add all code 11-7... |
10 11 12 |
a = hc(dataset, score='bde') plot(a) |
94c6dd68c 11-29-2016 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 |
from <- a$arcs[,1] to <- a$arcs[,2] m <- ncol(dataset) n <- nrow(dataset) #a function that return the variable position position <- function(x, data) { result <- which(x == names(data)) return(result) } from <- sapply(from,position,data=dataset) to <- sapply(to,position,data=dataset) score(a,data=dataset,type="bde") #############simple test sample############ test <- learning.test[1:5,1:3] test[2,3] <- "b" test <- data.frame(apply(test,2,as.character)) dag <- hc(test, score = "bic") score(dag,data=test,type="bde") #-10.37759 #bde score by hand score <- log(gamma(.5)/gamma(3.5)) + log(gamma(3.25)/gamma(.25)) + log(gamma(.25)/gamma(.25)) + log(gamma(.5)/gamma(2.5)) + log(gamma(1.25)/gamma(.25)) + log(gamma(1.25)/gamma(.25)) + log(gamma(.5)/gamma(3.5)) + log(gamma(2.25)/gamma(.25)) + log(gamma(1.25)/gamma(.25)) + log(gamma(.5)/gamma(2.5)) + log(gamma(.25)/gamma(.25)) + log(gamma(2.25)/gamma(.25)) write.csv(test,"./code/test.csv",row.names = F) ################################################ ######check all the graph of 3 variables######## ################################################ ######generate data set.seed(123) A <- sample(1:2,5000,replace = T) set.seed(428) B <- sample(1:2,5000,replace = T) set.seed(403) C <- sample(1:2,5000,replace = T) dataset <- data.frame(A,B,C) #save the dataset write.csv(dataset,"test_three_variables.csv",row.names = F) dataset <- read.csv("test_three_variables.csv") #make the variable as factor dataset <- apply(dataset,2,as.factor) dataset <- as.data.frame(dataset) score_bde <- numeric(25) #graph 1 A B C g1 <- empty.graph(names(dataset)) score_bde[1] <- score(g1,data = dataset, type="bde") #-10405.17 #graph 2 A->B C g2 <- set.arc(g1,"A","B") score_bde[2] <- score(g2,data = dataset, type="bde") #-10408.04 #graph 3 C->A->B g3 <- set.arc(g2,"C","A") score_bde[3] <- score(g3,data = dataset, type="bde") #-10411.15 #graph 4 B->A<-C g4 <- empty.graph(names(dataset)) modelstring(g4) <- "[B][C][A|B:C]" score_bde[4] <- score(g4,data = dataset, type="bde") #-10414.47 #graph 5 B->C->A g5 <- empty.graph(names(dataset)) modelstring(g5) <- "[B][C|B][A|C]" score_bde[5] <- score(g5,data = dataset, type="bde") #-10411.13 #graph 6 g6 <- empty.graph(names(dataset)) modelstring(g6) <- "[B][C|B][A|C:B]" score_bde[6] <- score(g6,data = dataset, type="bde") #-10417.33 #graph 7 g7 <- empty.graph(names(dataset)) modelstring(g7) <- "[C][A|C][B|C]" score_bde[7] <- score(g7,data = dataset, type="bde") #-10411.13 #graph 8 g8 <- empty.graph(names(dataset)) modelstring(g8) <- "[C][A|C][B|C:A]" score_bde[8] <- score(g8,data = dataset, type="bde") #-10417.33 #graph 9 g9 <- empty.graph(names(dataset)) modelstring(g9) <- "[C][A|C:B][B|C]" score_bde[9] <- score(g9,data = dataset, type="bde") #-10417.33 #graph 10 g10 <- empty.graph(names(dataset)) modelstring(g10) <- "[C|A][A][B]" score_bde[10] <- score(g10,data = dataset, type="bde") #-10408.27 #graph 11 g11 <- empty.graph(names(dataset)) modelstring(g11) <- "[C|A][A][B|A]" score_bde[11] <- score(g11,data = dataset, type="bde") #-10411.15 #graph 12 g12 <- empty.graph(names(dataset)) modelstring(g12) <- "[A|B][B][C|A]" score_bde[12] <- score(g12,data = dataset, type="bde") #-10411.15 #graph 13 g13 <- empty.graph(names(dataset)) modelstring(g13) <- "[A|B][B][C]" score_bde[13] <- score(g13,data = dataset, type="bde") #-10408.04 #graph 14 g14 <- empty.graph(names(dataset)) modelstring(g14) <- "[A][B][C|A:B]" score_bde[14] <- score(g14,data = dataset, type="bde") #-10414.45 #graph 15 g15 <- empty.graph(names(dataset)) modelstring(g15) <- "[A][B|A][C|A:B]" score_bde[15] <- score(g15,data = dataset, type="bde") #-10417.33 #graph 16 g16 <- empty.graph(names(dataset)) modelstring(g16) <- "[A|B][B][C|A:B]" score_bde[16] <- score(g16,data = dataset, type="bde") #-10417.33 #graph 17 g17 <- empty.graph(names(dataset)) modelstring(g17) <- "[A][B|C][C|A]" score_bde[17] <- score(g17,data = dataset, type="bde") #-10411.13 #graph 18 g18 <- empty.graph(names(dataset)) modelstring(g18) <- "[A][B|C:A][C|A]" score_bde[18] <- score(g18,data = dataset, type="bde") #-10417.33 #graph 19 g19 <- empty.graph(names(dataset)) modelstring(g19) <- "[A][B][C|B]" score_bde[19] <- score(g19,data = dataset, type="bde") #-10408.03 #graph 20 g20 <- empty.graph(names(dataset)) modelstring(g20) <- "[A][B|A][C|B]" score_bde[20] <- score(g20,data = dataset, type="bde") #-10410.9 #graph 21 g21 <- empty.graph(names(dataset)) modelstring(g21) <- "[A|B][B][C|B]" score_bde[21] <- score(g21,data = dataset, type="bde") #-10410.9 #graph 22 g22 <- empty.graph(names(dataset)) modelstring(g22) <- "[A][B|C][C]" score_bde[22] <- score(g22,data = dataset, type="bde") #-10408.03 #graph 23 g23 <- empty.graph(names(dataset)) modelstring(g23) <- "[A][B|C:A][C]" score_bde[23] <- score(g23,data = dataset, type="bde") #-10414.22 #graph 24 g24 <- empty.graph(names(dataset)) modelstring(g24) <- "[A|B][B|C][C]" score_bde[24] <- score(g24,data = dataset, type="bde") #-10410.9 #graph 25 g25 <- empty.graph(names(dataset)) modelstring(g25) <- "[A|C][B][C]" score_bde[25] <- score(g25,data = dataset, type="bde") #-10408.27 m <- ncol(dataset) n <- nrow(dataset) #a function that return the variable position position <- function(x, data) { result <- which(x == names(data)) return(result) } #bde score from python score_bde_python <- c(-10409.216534618594, -10413.72828956955, -10418.467521622148, -10423.856345541608, -10418.448531487986, -10418.44853148799, -10418.44853148799, -10428.34911035841, -10428.34911035841, -10413.95576667119, -10418.467521622146, -10418.467521622146, -10413.72828956955, -10423.837355407451, -10428.349110358407, -10428.349110358407, -10418.448531487988, -10428.349110358407, -10413.709299435392, -10418.221054386348, -10418.221054386348, -10413.709299435392, -10423.609878305811, -10418.221054386351, -10413.955766671188) arc <- data.frame() #output a pdf file pdf("plot.pdf") for (i in 1:25) { s <- paste0("g",i) tmp <- eval(parse(text=s)) #plot plot(tmp,main=paste("BDe score (python): ", score_bde_python[i], " BDe score (r): ", score_bde[i])) } dev.off() score_bde - score_bde_python for (i in 1:25) { s <- paste0("g",i) tmp <- eval(parse(text=s)) #plot plot(tmp,main=paste("BDe score: ", score_bde_python[i])) from <- tmp$arcs[,1] to <- tmp$arcs[,2] if (length(from) == 0) next else { from <- sapply(from,position,data=dataset) - 1 to <- sapply(to,position,data=dataset) - 1 l <- rep(i,length(from)) t <- data.frame(from, to, l) arc <- rbind(arc, t) } } names(arc) <- c("from","to","graph") write.csv(arc,"arc.csv",row.names = F) |
8375af764 add all code 11-7... |
281 |
|
8375af764 add all code 11-7... |
282 |