R: Matrices of Basic Probability Assignment Objects

bpamat {MuViCP}

R Documentation

Matrices of Basic Probability Assignment Objects

Description

These functions enhance the functionality provided via bpa objects. They essentially provide for the storage of several bpa objects at once as a matrix.

Usage

bpamat(info = NULL, mat = NULL)
## S3 method for class 'bpamat'
print(x, ...)

Arguments

`info`	A piece of auxiliary information that you want stored along with the matrix of bpa's. (Used internally to store the random seed, or the special character `'C'` if the bpa object resulted from a combination.)
`mat`	The matrix of bpa's. Each column represents a point, and the rows represent classes. Should have column names set to the row names of the points, and row names set to the names of the classes
`x`	The bpamat object to be printed.
`...`	Additional arguments to print method. Not Used.

Details

The ensemble function returns objects of this type.

Value

The bpamat function returns a list of functions which can be used to query and / or manipulate the create bpa object.

`set.info`	Takes a single argument which is set as the auxiliary information you want stored with the matrix
`get.info`	Returns the auxiliary information stored with the matrix
`assign.mat`	Takes a single argument, which should be a matrix of bpa's, to be stored inside the bpamat object.
`get.classify`	Returns a vector as long as the number of points stored in the `bpamat` object. The elements are named after the points, and are the current classification of the point, based on the bpamat object.
`get.point`	Returns the bpa corresponding to a single point, whose name is passed as the argument.
`get.mat`	Returns the matrix of bpa's.
`get.setlist`	Returns the class names that occur in the current matrix
`get.pointlist`	Returns the names of all the points whose bpa's are stored in the current matrix.

Author(s)

Mohit Dayal

Examples

data(cancer)
table(cancer$V2)
colnames(cancer)[1:2] <- c('id', 'type')

cancer.d <- as.matrix(cancer[,3:32])
labs <- cancer$type
test_size <- floor(0.15*nrow(cancer.d))
train <- sample(1:nrow(cancer.d), size = nrow(cancer.d) - test_size)
test <- which(!(1:569 %in% train))
truelabs <- labs[test]

projectron <- function(A)
    cancer.d %*% A

kdebel <- kde_bel.builder(labs = labs[train], test = test, train =
train)

##A projection
seed1 <- .Random.seed
F1 <- projectron(basis_random(30))
x1 <- kdebel(F1)
y1 <- bpamat(info = seed1, mat = x1)
y1
predicted1 <- y1$get.classify()
table(truelabs, predicted1)

##Another projection
seed2 <- .Random.seed
F2 <- projectron(basis_random(30))
x2 <- kdebel(F2)
y2 <- bpamat(info = seed2, mat = x2)
y2
predicted2 <- y2$get.classify()
table(truelabs, predicted2)

z1 <- combine.bpamat.bs(y1, y2)
z2 <- combine.bpamat.ds(y1, y2)
table(truelabs, z1$get.classify())
table(truelabs, z2$get.classify())

##Same result
w1 <- combine.bpamat.list.bs(list(y1, y2))
w2 <- combine.bpamat.list.ds(list(y1, y2))