[epclust.git] / epclust / tests / testthat / helper.clustering.R

# Shorthand: map 1->1, 2->2, 3->3, 4->1, ..., 149->2, 150->3, ... (is base==3)
I = function(i, base)
	(i-1) %% base + 1

# Compute the sum of (normalized) sum of squares of closest distances to a medoid.
# Note: medoids can be a big.matrix
computeDistortion = function(series, medoids)
{
	if (bigmemory::is.big.matrix(medoids))
		medoids = medoids[,] #extract standard matrix

	n = nrow(series) ; L = ncol(series)
	distortion = 0.
	for (i in seq_len(n))
		distortion = distortion + min( colSums( sweep(medoids,1,series[,i],'-')^2 ) / L )

	distortion / n
}
Commit	Line	Data
d9bb53c5	1	# Shorthand: map 1->1, 2->2, 3->3, 4->1, ..., 149->2, 150->3, ... (is base==3)
0486fbad BA	2	I = function(i, base)
	3	(i-1) %% base + 1
	4
d9bb53c5 BA	5	# Compute the sum of (normalized) sum of squares of closest distances to a medoid.
d9bb53c5 BA	6	# Note: medoids can be a big.matrix
0486fbad BA	7	computeDistortion = function(series, medoids)
0486fbad BA	8	{
d9bb53c5 BA	9	if (bigmemory::is.big.matrix(medoids))
	10	medoids = medoids[,] #extract standard matrix
	11
0486fbad BA	12	n = nrow(series) ; L = ncol(series)
0486fbad BA	13	distortion = 0.
0486fbad BA	14	for (i in seq_len(n))
0486fbad BA	15	distortion = distortion + min( colSums( sweep(medoids,1,series[,i],'-')^2 ) / L )
d9bb53c5	16
0486fbad BA	17	distortion / n
0486fbad BA	18	}