Weight and Fuse Objects

Weight and fuse distance matrices based on the relative weights.

FuseData(..., project_k = 10, zero_percent = 0.7)

Arguments

...

FuseNet objects to fuse.

project_k

Number of nearest neighbors to project based on geometric sketches, if it has been run. Default is 10.

zero_percent

Zero-entry percentage threshold. If the number of zero entries in the returned matrices is above this number, a sparse matrix will be returned. Default is 0.7 aka 70%.

Value

Returns a list with entries:

• fused_weight, n x M matrix with n rows of objects or modality to fuse and M columns of data points.

• fused_dist, M x M fused distance matrix.

Examples

{
object1 <- InitiateFuseNet(t(iris[,1:2]), project_name = "FuseNet", k = 3)
object1 <- RunFuseNet(object1, n_iters = 1, k = 3, ratio = 0.5, n_cores = 1)
object2 <- InitiateFuseNet(t(iris[,3:4]), project_name = "FuseNet", k = 3)
object2 <- RunFuseNet(object2, n_iters = 2, k = 3, ratio = 0.5, n_cores = 1)
fused.data <- FuseData(object1, object2)
}
#> Initiate FuseNet
#> Normalize Data
#> Find Nearest Neighbors
#> Matrix Power
#> Finalize
#> Run Bootstrapping
#> Finalize
#> Initiate FuseNet
#> Normalize Data
#> Find Nearest Neighbors
#> Matrix Power
#> Finalize
#> Run Bootstrapping
#> Finalize