The possible weights to be considered is a sequence from 0 (favour
auxiliary data) to 1 (favour primary data). Each possible
combination of weights for nclass
classes must be
tested. The thetas
function produces a weight matrix
for nclass
columns (one for each class) with all possible
weight combinations (number of rows).
Examples
dim(thetas(4, by = 0.5))
#> Weigths:
#> (0, 0.5, 1)
#> [1] 81 4
dim(thetas(4, by = 0.2))
#> Weigths:
#> (0, 0.2, 0.4, 0.6, 0.8, 1)
#> [1] 1296 4
dim(thetas(5, by = 0.2))
#> Weigths:
#> (0, 0.2, 0.4, 0.6, 0.8, 1)
#> [1] 7776 5
dim(thetas(5, length.out = 5))
#> Weigths:
#> (0, 0.25, 0.5, 0.75, 1)
#> [1] 3125 5
dim(thetas(6, by = 0.2))
#> Weigths:
#> (0, 0.2, 0.4, 0.6, 0.8, 1)
#> [1] 46656 6