Description Usage Arguments Value Author(s) See Also Examples

View source: R/predict.binaryGP.R

The function computes the predicted response and its variance as well as its confidence interval.

1 2 3 |

`object` |
a class binaryGP object estimated by |

`xnew` |
a testing matrix with dimension |

`conf.level` |
a value from 0 to 1 specifying the level of confidence interval. The default is 0.95. |

`sim.number` |
a positive integer specifying the simulation number for Monte-Carlo method. The default is 101. |

`...` |
for compatibility with generic method |

`mean` |
a matrix with dimension |

`var` |
a matrix with dimension |

`upper.bound` |
a matrix with dimension |

`lower.bound` |
a matrix with dimension |

`y_pred` |
a matrix with dimension |

Chih-Li Sung <[email protected]>

`binaryGP_fit`

for estimation of the binary Gaussian process.

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 | ```
library(binaryGP)
##### Testing function: cos(x1 + x2) * exp(x1*x2) with TT sequences #####
##### Thanks to Sonja Surjanovic and Derek Bingham, Simon Fraser University #####
test_function <- function(X, TT)
{
x1 <- X[,1]
x2 <- X[,2]
eta_1 <- cos(x1 + x2) * exp(x1*x2)
p_1 <- exp(eta_1)/(1+exp(eta_1))
y_1 <- rep(NA, length(p_1))
for(i in 1:length(p_1)) y_1[i] <- rbinom(1,1,p_1[i])
Y <- y_1
P <- p_1
if(TT > 1){
for(tt in 2:TT){
eta_2 <- 0.3 * y_1 + eta_1
p_2 <- exp(eta_2)/(1+exp(eta_2))
y_2 <- rep(NA, length(p_2))
for(i in 1:length(p_2)) y_2[i] <- rbinom(1,1,p_2[i])
Y <- cbind(Y, y_2)
P <- cbind(P, p_2)
y_1 <- y_2
}
}
return(list(Y = Y, P = P))
}
set.seed(1)
n <- 30
n.test <- 10
d <- 2
X <- matrix(runif(d * n), ncol = d)
X.test <- matrix(runif(d * n.test), ncol = d)
##### without time-series #####
Y <- test_function(X, 1)$Y ## Y is a vector
test.out <- test_function(X.test, 1)
Y.test <- test.out$Y
P.true <- test.out$P
# fitting
binaryGP.model <- binaryGP_fit(X = X, Y = Y)
# prediction
binaryGP.prediction <- predict(binaryGP.model, xnew = X.test)
print(binaryGP.prediction$mean)
print(binaryGP.prediction$var)
print(binaryGP.prediction$upper.bound)
print(binaryGP.prediction$lower.bound)
##### with time-series #####
Y <- test_function(X, 10)$Y ## Y is a matrix with 10 columns
test.out <- test_function(X.test, 10)
Y.test <- test.out$Y
P.true <- test.out$P
# fitting
binaryGP.model <- binaryGP_fit(X = X, Y = Y, R = 1)
# prediction
binaryGP.prediction <- predict(binaryGP.model, xnew = X.test)
print(binaryGP.prediction$mean)
print(binaryGP.prediction$var)
print(binaryGP.prediction$upper.bound)
print(binaryGP.prediction$lower.bound)
``` |

binaryGP documentation built on Sept. 19, 2017, 9:02 a.m.

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.