tcR (version 2.2.4)

ozScore: Overlap Z-score.

Description

Compute OZ-scores ("overlap Z scores") for values in the given matrix of overlaps, i.e.,. for each value compute the number of standart deviations from the mean of the matrix.

Usage

ozScore(.mat, .symm = T, .as.matrix = F, .val.col = c("norm", "abs",
  "oz"))

Arguments

.mat

Matrix with overlap values.

.symm

If T then remove lower triangle matrix from counting. Doesn't work if the matrix has different number of rows and columns.

.as.matrix

If T then return

.val.col

If .as.matrix is T then this is a name of the column to build matrix upon: either "oz" for the OZ-score column, "abs" for the absolute OZ-score column, or "norm" for the normalised absolute OZ-score column.

See Also

repOverlap, intersectClonesets, permutDistTest

Examples

# NOT RUN {
data(twb)
mat <- repOverlap(twb)
ozScore(mat)
# Take 3x3 matrix
ozScore(mat[1:3, 1:3])
# Return as matrix with OZ scores
ozmat <- ozScore(mat, T, T, 'oz')
# Return as matrix with normalised absolute OZ scores
oznorm <- ozScore(mat, T, T, 'norm')
# Plot it as boxplots
sb <- matrixSubgroups(oznorm, list(tw1 = c('Subj.A', 'Subj.B'), tw2 = c('Subj.C', 'Subj.D')));
vis.group.boxplot(sb)
# }