Generate random data within specified archetypes

generate_arc() generates the matrix of archetypes using specified coordinates (some noise added).

generate_data() produces matrix of random data that fits a polytope defined by archetypes by multiplying position of archetypes by random weigths (that sum to 1)

generate_arc(arc_coord = list(c(5, 0), c(-10, 15), c(-30, -20)),
  mean = 0, sd = 1)

generate_data(archetypes, N_examples = 10000, jiiter = 0.1, size = 1)

Arguments

arc_coord	list of archetype coordinates, one numeric vector (length of N dimensions) per each archetype.
mean	mean of random distribution added to arc_coord
sd	standard deviationn of random distribution added to arc_coord
archetypes	matrix of archetypes of dim(dimensions, archetypes)
N_examples	number of examples to be generated
jiiter	add noise to weigth so that data is not a perfect polytope (e.g. triangle, see examples)
size	scale the data within a polytope

Value

generate_arc() object of class "random_arc" (similar to "pch_fit"), element XC is a matrix of archetypes of dim(dimensions, archetypes)

generate_data() matrix of archetypes of dim(dimensions, examples)

See also

fit_pch, arch_dist

Examples

# Random data that fits into the triangle
set.seed(4355)
archetypes = generate_arc(arc_coord = list(c(5, 0), c(-10, 15), c(-30, -20)),
                          mean = 0, sd = 1)
data = generate_data(archetypes$XC, N_examples = 1e4, jiiter = 0.04, size = 0.9)
# Find Euclidian distance between data points and archetypes
distance = arch_dist(data, archetypes)
#> Warning: first element used of 'length.out' argument
#> Error in seq_len(ncol(archetypes)): argument must be coercible to non-negative integer
# Find Euclidian distance between archetypes
arc_distance = arch_dist(archetypes, archetypes)
#> Warning: first element used of 'length.out' argument
#> Error in seq_len(ncol(archetypes)): argument must be coercible to non-negative integer