R/biips-object.r
A biips object is returned by function biips_model. It
represents a Bayesian graphical model described in BUGS language.
The method biips_build_sampler assigns a sampler to each node
of the graph. In order to specify the proposal used by the SMC
algorithm, this function has to be called before biips_smc_samples.
Otherwise, it will be automatically called by biips_smc_samples
with the default parameters.
is.biips(object) # S3 method for biips print(x, ...) # S3 method for biips biips_variable_names(object, ...) # S3 method for biips variable.names(object, ...) # S3 method for biips biips_nodes(object, type, observed, ...) # S3 method for biips biips_print_dot(object, file, ...) # S3 method for biips biips_build_sampler(object, proposal = "auto", ...)
| object, x |
|
|---|---|
| ... | Additional arguments to be passed to default methods. |
| type | string. Return only a specific type of node. Possible values
are |
| observed | logical. Return only observed or unobserved nodes. Default returns all. |
| file | string. Path of the output file. |
| proposal | string. The type of proposal used by the SMC algorithm.
Possible values are |
The function is.biips returns TRUE if the object is of
class biips.
The method biips_variable_names returns a character vector.
Names of node arrays used in the model.
The method variable.names is an alias for
biips_variable_names.
The method biips_nodes returns a data.frame with a row
for each node of the graphical model sorted in a topological order with the
following columns:
integer. node ids.
string. node names.
string. node types ('const', 'logic' or
'stoch').
logical. TRUE if the node is observed.
logical. TRUE if the node is discrete.
integer. node sampling iteration of unobserved nodes,
starting at 0. NA if the node if observed.
string. node sampler name for stochastic unobserved nodes. An empty string for other types of nodes.
modelfile <- system.file('extdata', 'hmm.bug', package = 'rbiips') stopifnot(nchar(modelfile) > 0) cat(readLines(modelfile), sep = '\n')#> var c_true[tmax], x_true[tmax], c[tmax], x[tmax], y[tmax] #> #> data #> { #> x_true[1] ~ dnorm(0, 1/5) #> y[1] ~ dnorm(x_true[1], exp(logtau_true)) #> for (t in 2:tmax) #> { #> c_true[t] ~ dcat(p) #> x_true[t] ~ dnorm(0.5*x_true[t-1]+25*x_true[t-1]/(1+x_true[t-1]^2)+8*cos(1.2*(t-1)), ifelse(c_true[t]==1, 1/10, 1/100)) #> y[t] ~ dnorm(x_true[t]/4, exp(logtau_true)) #> } #> } #> #> model #> { #> logtau ~ dunif(-3, 3) #> x[1] ~ dnorm(0, 1/5) #> y[1] ~ dnorm(x[1], exp(logtau)) #> for (t in 2:tmax) #> { #> c[t] ~ dcat(p) #> x[t] ~ dnorm(0.5*x[t-1]+25*x[t-1]/(1+x[t-1]^2)+8*cos(1.2*(t-1)), ifelse(c[t]==1, 1/10, 1/100)) #> y[t] ~ dnorm(x[t]/4, exp(logtau)) #> } #> }data <- list(tmax = 10, p = c(.5, .5), logtau_true = log(1), logtau = log(1)) model <- biips_model(modelfile, data, sample_data = TRUE)#> * Parsing model in: /home/adrien/Dropbox/workspace/rbiips/inst/extdata/hmm.bug #> * Compiling data graph #> Declaring variables #> Resolving undeclared variables #> Allocating nodes #> Graph size: 169 #> Sampling data #> Reading data back into data table #> * Compiling model graph #> Declaring variables #> Resolving undeclared variables #> Allocating nodes #> Graph size: 180#\dontrun{ #tmax <- 10 #p <- c(.5, .5) #logtau_true <- log(1) #logtau <- logtau_true #datanames <- c('tmax', 'p', 'logtau_true', 'logtau') #model <- biips_model(modelfile, datanames, sample_data = TRUE) #} is.biips(model)#> [1] TRUEprint(model)#> Biips model: #> #> var c_true[tmax], x_true[tmax], c[tmax], x[tmax], y[tmax] #> #> data #> { #> x_true[1] ~ dnorm(0, 1/5) #> y[1] ~ dnorm(x_true[1], exp(logtau_true)) #> for (t in 2:tmax) #> { #> c_true[t] ~ dcat(p) #> x_true[t] ~ dnorm(0.5*x_true[t-1]+25*x_true[t-1]/(1+x_true[t-1]^2)+8*cos(1.2*(t-1)), ifelse(c_true[t]==1, 1/10, 1/100)) #> y[t] ~ dnorm(x_true[t]/4, exp(logtau_true)) #> } #> } #> #> model #> { #> logtau ~ dunif(-3, 3) #> x[1] ~ dnorm(0, 1/5) #> y[1] ~ dnorm(x[1], exp(logtau)) #> for (t in 2:tmax) #> { #> c[t] ~ dcat(p) #> x[t] ~ dnorm(0.5*x[t-1]+25*x[t-1]/(1+x[t-1]^2)+8*cos(1.2*(t-1)), ifelse(c[t]==1, 1/10, 1/100)) #> y[t] ~ dnorm(x[t]/4, exp(logtau)) #> } #> } #> #> #> Fully observed variables: #> logtau logtau_true p tmax x_true y #> #> Partially observed variables: #> c_truemodel$data()#> $c_true #> [1] NA 1 2 1 1 2 1 2 2 1 #> #> $logtau #> [1] 0 #> #> $logtau_true #> [1] 0 #> #> $p #> [1] 0.5 0.5 #> #> $tmax #> [1] 10 #> #> $x_true #> [1] -0.1080991 2.6227064 3.5849787 3.7873451 10.2074744 27.5563233 #> [7] 18.5383870 20.5129987 14.4393068 9.4665880 #> #> $y #> [1] -0.10944153 1.76192206 -0.99069457 0.91096533 1.64333996 6.31411620 #> [7] 5.13625227 6.88850807 3.77524079 0.05766708 #>variable.names(model)#> [1] "c" "c_true" "logtau" "logtau_true" "p" #> [6] "tmax" "x" "x_true" "y"biips_variable_names(model)#> [1] "c" "c_true" "logtau" "logtau_true" "p" #> [6] "tmax" "x" "x_true" "y"biips_nodes(model)#> id name type #> 1 0 c_true[2] const #> 2 1 c_true[3] const #> 3 2 c_true[4] const #> 4 3 c_true[5] const #> 5 4 c_true[6] const #> 6 5 c_true[7] const #> 7 6 c_true[8] const #> 8 7 c_true[9] const #> 9 8 c_true[10] const #> 10 9 logtau_true const #> 11 10 p const #> 12 11 tmax const #> 13 12 x_true const #> 14 13 3 const #> 15 14 (-3) logic #> 16 15 logtau stoch #> 17 21 (exp(logtau)) logic #> 18 16 0 const #> 19 17 1 const #> 20 55 (3-1) logic #> 21 18 5 const #> 22 19 (1/5) logic #> 23 87 (5-1) logic #> 24 20 x[1] stoch #> 25 22 y[1] stoch #> 26 23 c[2] stoch #> 27 39 (c[2]==1) logic #> 28 24 0.5 const #> 29 25 (0.5*x[1]) logic #> 30 26 25 const #> 31 27 (25*x[1]) logic #> 32 28 2 const #> 33 29 (x[1]^2) logic #> 34 30 (1+(x[1]^2)) logic #> 35 31 ((25*x[1])/(1+(x[1]^2))) logic #> 36 34 (2-1) logic #> 37 32 8 const #> 38 137 (8-1) logic #> 39 33 1.2 const #> 40 35 (1.2*(2-1)) logic #> 41 36 (cos((1.2*(2-1)))) logic #> 42 37 (8*(cos((1.2*(2-1))))) logic #> 43 38 ((0.5*x[1])+((25*x[1])/(1+(x[1]^2)))+(8*(cos((1.2*(2-1)))))) logic #> 44 56 (1.2*(3-1)) logic #> 45 57 (cos((1.2*(3-1)))) logic #> 46 58 (8*(cos((1.2*(3-1))))) logic #> 47 88 (1.2*(5-1)) logic #> 48 89 (cos((1.2*(5-1)))) logic #> 49 90 (8*(cos((1.2*(5-1))))) logic #> 50 138 (1.2*(8-1)) logic #> 51 139 (cos((1.2*(8-1)))) logic #> 52 140 (8*(cos((1.2*(8-1))))) logic #> 53 40 10 const #> 54 41 (1/10) logic #> 55 170 (10-1) logic #> 56 171 (1.2*(10-1)) logic #> 57 172 (cos((1.2*(10-1)))) logic #> 58 173 (8*(cos((1.2*(10-1))))) logic #> 59 42 100 const #> 60 43 (1/100) logic #> 61 44 (ifelse((c[2]==1),(1/10),(1/100))) logic #> 62 45 x[2] stoch #> 63 50 (0.5*x[2]) logic #> 64 51 (25*x[2]) logic #> 65 52 (x[2]^2) logic #> 66 53 (1+(x[2]^2)) logic #> 67 54 ((25*x[2])/(1+(x[2]^2))) logic #> 68 59 ((0.5*x[2])+((25*x[2])/(1+(x[2]^2)))+(8*(cos((1.2*(3-1)))))) logic #> 69 46 4 const #> 70 47 (x[2]/4) logic #> 71 48 y[2] stoch #> 72 71 (4-1) logic #> 73 72 (1.2*(4-1)) logic #> 74 73 (cos((1.2*(4-1)))) logic #> 75 74 (8*(cos((1.2*(4-1))))) logic #> 76 49 c[3] stoch #> 77 60 (c[3]==1) logic #> 78 61 (ifelse((c[3]==1),(1/10),(1/100))) logic #> 79 62 x[3] stoch #> 80 63 (x[3]/4) logic #> 81 64 y[3] stoch #> 82 66 (0.5*x[3]) logic #> 83 67 (25*x[3]) logic #> 84 68 (x[3]^2) logic #> 85 69 (1+(x[3]^2)) logic #> 86 70 ((25*x[3])/(1+(x[3]^2))) logic #> 87 75 ((0.5*x[3])+((25*x[3])/(1+(x[3]^2)))+(8*(cos((1.2*(4-1)))))) logic #> 88 65 c[4] stoch #> 89 76 (c[4]==1) logic #> 90 77 (ifelse((c[4]==1),(1/10),(1/100))) logic #> 91 78 x[4] stoch #> 92 79 (x[4]/4) logic #> 93 80 y[4] stoch #> 94 82 (0.5*x[4]) logic #> 95 83 (25*x[4]) logic #> 96 84 (x[4]^2) logic #> 97 85 (1+(x[4]^2)) logic #> 98 86 ((25*x[4])/(1+(x[4]^2))) logic #> 99 91 ((0.5*x[4])+((25*x[4])/(1+(x[4]^2)))+(8*(cos((1.2*(5-1)))))) logic #> 100 81 c[5] stoch #> 101 92 (c[5]==1) logic #> 102 93 (ifelse((c[5]==1),(1/10),(1/100))) logic #> 103 94 x[5] stoch #> 104 95 (x[5]/4) logic #> 105 96 y[5] stoch #> 106 98 (0.5*x[5]) logic #> 107 99 (25*x[5]) logic #> 108 100 (x[5]^2) logic #> 109 101 (1+(x[5]^2)) logic #> 110 102 ((25*x[5])/(1+(x[5]^2))) logic #> 111 97 c[6] stoch #> 112 109 (c[6]==1) logic #> 113 110 (ifelse((c[6]==1),(1/10),(1/100))) logic #> 114 103 6 const #> 115 104 (6-1) logic #> 116 105 (1.2*(6-1)) logic #> 117 106 (cos((1.2*(6-1)))) logic #> 118 107 (8*(cos((1.2*(6-1))))) logic #> 119 108 ((0.5*x[5])+((25*x[5])/(1+(x[5]^2)))+(8*(cos((1.2*(6-1)))))) logic #> 120 111 x[6] stoch #> 121 112 (x[6]/4) logic #> 122 113 y[6] stoch #> 123 115 (0.5*x[6]) logic #> 124 116 (25*x[6]) logic #> 125 117 (x[6]^2) logic #> 126 118 (1+(x[6]^2)) logic #> 127 119 ((25*x[6])/(1+(x[6]^2))) logic #> 128 114 c[7] stoch #> 129 126 (c[7]==1) logic #> 130 127 (ifelse((c[7]==1),(1/10),(1/100))) logic #> 131 120 7 const #> 132 121 (7-1) logic #> 133 122 (1.2*(7-1)) logic #> 134 123 (cos((1.2*(7-1)))) logic #> 135 124 (8*(cos((1.2*(7-1))))) logic #> 136 125 ((0.5*x[6])+((25*x[6])/(1+(x[6]^2)))+(8*(cos((1.2*(7-1)))))) logic #> 137 128 x[7] stoch #> 138 129 (x[7]/4) logic #> 139 130 y[7] stoch #> 140 132 (0.5*x[7]) logic #> 141 133 (25*x[7]) logic #> 142 134 (x[7]^2) logic #> 143 135 (1+(x[7]^2)) logic #> 144 136 ((25*x[7])/(1+(x[7]^2))) logic #> 145 141 ((0.5*x[7])+((25*x[7])/(1+(x[7]^2)))+(8*(cos((1.2*(8-1)))))) logic #> 146 131 c[8] stoch #> 147 142 (c[8]==1) logic #> 148 143 (ifelse((c[8]==1),(1/10),(1/100))) logic #> 149 144 x[8] stoch #> 150 145 (x[8]/4) logic #> 151 146 y[8] stoch #> 152 148 (0.5*x[8]) logic #> 153 149 (25*x[8]) logic #> 154 150 (x[8]^2) logic #> 155 151 (1+(x[8]^2)) logic #> 156 152 ((25*x[8])/(1+(x[8]^2))) logic #> 157 147 c[9] stoch #> 158 159 (c[9]==1) logic #> 159 160 (ifelse((c[9]==1),(1/10),(1/100))) logic #> 160 153 9 const #> 161 154 (9-1) logic #> 162 155 (1.2*(9-1)) logic #> 163 156 (cos((1.2*(9-1)))) logic #> 164 157 (8*(cos((1.2*(9-1))))) logic #> 165 158 ((0.5*x[8])+((25*x[8])/(1+(x[8]^2)))+(8*(cos((1.2*(9-1)))))) logic #> 166 161 x[9] stoch #> 167 162 (x[9]/4) logic #> 168 163 y[9] stoch #> 169 165 (0.5*x[9]) logic #> 170 166 (25*x[9]) logic #> 171 167 (x[9]^2) logic #> 172 168 (1+(x[9]^2)) logic #> 173 169 ((25*x[9])/(1+(x[9]^2))) logic #> 174 174 ((0.5*x[9])+((25*x[9])/(1+(x[9]^2)))+(8*(cos((1.2*(10-1)))))) logic #> 175 164 c[10] stoch #> 176 175 (c[10]==1) logic #> 177 176 (ifelse((c[10]==1),(1/10),(1/100))) logic #> 178 177 x[10] stoch #> 179 178 (x[10]/4) logic #> 180 179 y[10] stoch #> observed discrete #> 1 TRUE TRUE #> 2 TRUE TRUE #> 3 TRUE TRUE #> 4 TRUE TRUE #> 5 TRUE TRUE #> 6 TRUE TRUE #> 7 TRUE TRUE #> 8 TRUE TRUE #> 9 TRUE TRUE #> 10 TRUE TRUE #> 11 TRUE FALSE #> 12 TRUE TRUE #> 13 TRUE FALSE #> 14 TRUE TRUE #> 15 TRUE TRUE #> 16 TRUE FALSE #> 17 TRUE FALSE #> 18 TRUE TRUE #> 19 TRUE TRUE #> 20 TRUE TRUE #> 21 TRUE TRUE #> 22 TRUE FALSE #> 23 TRUE TRUE #> 24 FALSE FALSE #> 25 TRUE FALSE #> 26 FALSE TRUE #> 27 FALSE TRUE #> 28 TRUE FALSE #> 29 FALSE FALSE #> 30 TRUE TRUE #> 31 FALSE FALSE #> 32 TRUE TRUE #> 33 FALSE FALSE #> 34 FALSE FALSE #> 35 FALSE FALSE #> 36 TRUE TRUE #> 37 TRUE TRUE #> 38 TRUE TRUE #> 39 TRUE FALSE #> 40 TRUE FALSE #> 41 TRUE FALSE #> 42 TRUE FALSE #> 43 FALSE FALSE #> 44 TRUE FALSE #> 45 TRUE FALSE #> 46 TRUE FALSE #> 47 TRUE FALSE #> 48 TRUE FALSE #> 49 TRUE FALSE #> 50 TRUE FALSE #> 51 TRUE FALSE #> 52 TRUE FALSE #> 53 TRUE TRUE #> 54 TRUE FALSE #> 55 TRUE TRUE #> 56 TRUE FALSE #> 57 TRUE FALSE #> 58 TRUE FALSE #> 59 TRUE TRUE #> 60 TRUE FALSE #> 61 FALSE FALSE #> 62 FALSE FALSE #> 63 FALSE FALSE #> 64 FALSE FALSE #> 65 FALSE FALSE #> 66 FALSE FALSE #> 67 FALSE FALSE #> 68 FALSE FALSE #> 69 TRUE TRUE #> 70 FALSE FALSE #> 71 TRUE FALSE #> 72 TRUE TRUE #> 73 TRUE FALSE #> 74 TRUE FALSE #> 75 TRUE FALSE #> 76 FALSE TRUE #> 77 FALSE TRUE #> 78 FALSE FALSE #> 79 FALSE FALSE #> 80 FALSE FALSE #> 81 TRUE FALSE #> 82 FALSE FALSE #> 83 FALSE FALSE #> 84 FALSE FALSE #> 85 FALSE FALSE #> 86 FALSE FALSE #> 87 FALSE FALSE #> 88 FALSE TRUE #> 89 FALSE TRUE #> 90 FALSE FALSE #> 91 FALSE FALSE #> 92 FALSE FALSE #> 93 TRUE FALSE #> 94 FALSE FALSE #> 95 FALSE FALSE #> 96 FALSE FALSE #> 97 FALSE FALSE #> 98 FALSE FALSE #> 99 FALSE FALSE #> 100 FALSE TRUE #> 101 FALSE TRUE #> 102 FALSE FALSE #> 103 FALSE FALSE #> 104 FALSE FALSE #> 105 TRUE FALSE #> 106 FALSE FALSE #> 107 FALSE FALSE #> 108 FALSE FALSE #> 109 FALSE FALSE #> 110 FALSE FALSE #> 111 FALSE TRUE #> 112 FALSE TRUE #> 113 FALSE FALSE #> 114 TRUE TRUE #> 115 TRUE TRUE #> 116 TRUE FALSE #> 117 TRUE FALSE #> 118 TRUE FALSE #> 119 FALSE FALSE #> 120 FALSE FALSE #> 121 FALSE FALSE #> 122 TRUE FALSE #> 123 FALSE FALSE #> 124 FALSE FALSE #> 125 FALSE FALSE #> 126 FALSE FALSE #> 127 FALSE FALSE #> 128 FALSE TRUE #> 129 FALSE TRUE #> 130 FALSE FALSE #> 131 TRUE TRUE #> 132 TRUE TRUE #> 133 TRUE FALSE #> 134 TRUE FALSE #> 135 TRUE FALSE #> 136 FALSE FALSE #> 137 FALSE FALSE #> 138 FALSE FALSE #> 139 TRUE FALSE #> 140 FALSE FALSE #> 141 FALSE FALSE #> 142 FALSE FALSE #> 143 FALSE FALSE #> 144 FALSE FALSE #> 145 FALSE FALSE #> 146 FALSE TRUE #> 147 FALSE TRUE #> 148 FALSE FALSE #> 149 FALSE FALSE #> 150 FALSE FALSE #> 151 TRUE FALSE #> 152 FALSE FALSE #> 153 FALSE FALSE #> 154 FALSE FALSE #> 155 FALSE FALSE #> 156 FALSE FALSE #> 157 FALSE TRUE #> 158 FALSE TRUE #> 159 FALSE FALSE #> 160 TRUE TRUE #> 161 TRUE TRUE #> 162 TRUE FALSE #> 163 TRUE FALSE #> 164 TRUE FALSE #> 165 FALSE FALSE #> 166 FALSE FALSE #> 167 FALSE FALSE #> 168 TRUE FALSE #> 169 FALSE FALSE #> 170 FALSE FALSE #> 171 FALSE FALSE #> 172 FALSE FALSE #> 173 FALSE FALSE #> 174 FALSE FALSE #> 175 FALSE TRUE #> 176 FALSE TRUE #> 177 FALSE FALSE #> 178 FALSE FALSE #> 179 FALSE FALSE #> 180 TRUE FALSE#\dontrun{ #dotfile <- 'hmm.dot' #biips_print_dot(model, dotfile) #cat(readLines(dotfile), sep = '\n') #} biips_build_sampler(model, proposal = 'prior')#> * Assigning node samplersbiips_nodes(model, type = 'stoch', observed = FALSE)#> id name type observed discrete iteration sampler #> 24 20 x[1] stoch FALSE FALSE 0 Prior #> 26 23 c[2] stoch FALSE TRUE 1 Prior #> 62 45 x[2] stoch FALSE FALSE 1 Prior #> 76 49 c[3] stoch FALSE TRUE 2 Prior #> 79 62 x[3] stoch FALSE FALSE 2 Prior #> 88 65 c[4] stoch FALSE TRUE 3 Prior #> 91 78 x[4] stoch FALSE FALSE 3 Prior #> 100 81 c[5] stoch FALSE TRUE 4 Prior #> 103 94 x[5] stoch FALSE FALSE 4 Prior #> 111 97 c[6] stoch FALSE TRUE 5 Prior #> 120 111 x[6] stoch FALSE FALSE 5 Prior #> 128 114 c[7] stoch FALSE TRUE 6 Prior #> 137 128 x[7] stoch FALSE FALSE 6 Prior #> 146 131 c[8] stoch FALSE TRUE 7 Prior #> 149 144 x[8] stoch FALSE FALSE 7 Prior #> 157 147 c[9] stoch FALSE TRUE 8 Prior #> 166 161 x[9] stoch FALSE FALSE 8 Prior #> 175 164 c[10] stoch FALSE TRUE 9 Prior #> 178 177 x[10] stoch FALSE FALSE 9 Priorbiips_build_sampler(model, proposal = 'auto')#> * Assigning node samplersbiips_nodes(model, type = 'stoch', observed = FALSE)#> id name type observed discrete iteration #> 24 20 x[1] stoch FALSE FALSE 0 #> 26 23 c[2] stoch FALSE TRUE 1 #> 62 45 x[2] stoch FALSE FALSE 1 #> 76 49 c[3] stoch FALSE TRUE 2 #> 79 62 x[3] stoch FALSE FALSE 2 #> 88 65 c[4] stoch FALSE TRUE 3 #> 91 78 x[4] stoch FALSE FALSE 3 #> 100 81 c[5] stoch FALSE TRUE 4 #> 103 94 x[5] stoch FALSE FALSE 4 #> 111 97 c[6] stoch FALSE TRUE 5 #> 120 111 x[6] stoch FALSE FALSE 5 #> 128 114 c[7] stoch FALSE TRUE 6 #> 137 128 x[7] stoch FALSE FALSE 6 #> 146 131 c[8] stoch FALSE TRUE 7 #> 149 144 x[8] stoch FALSE FALSE 7 #> 157 147 c[9] stoch FALSE TRUE 8 #> 166 161 x[9] stoch FALSE FALSE 8 #> 175 164 c[10] stoch FALSE TRUE 9 #> 178 177 x[10] stoch FALSE FALSE 9 #> sampler #> 24 ConjugateNormal_knownPrec #> 26 Prior #> 62 ConjugateNormal_knownPrec_linearMean #> 76 Prior #> 79 ConjugateNormal_knownPrec_linearMean #> 88 Prior #> 91 ConjugateNormal_knownPrec_linearMean #> 100 Prior #> 103 ConjugateNormal_knownPrec_linearMean #> 111 Prior #> 120 ConjugateNormal_knownPrec_linearMean #> 128 Prior #> 137 ConjugateNormal_knownPrec_linearMean #> 146 Prior #> 149 ConjugateNormal_knownPrec_linearMean #> 157 Prior #> 166 ConjugateNormal_knownPrec_linearMean #> 175 Prior #> 178 ConjugateNormal_knownPrec_linearMean