R/visualization.R

In antagomir/netresponse: Functional Network Analysis

# 'The language of science is the language of probability, and not of p-values.'
# -- Luis Pericchi

#' @title Set breaks
#' @description Set breakpoints for two-way color palette.
#' @usage set.breaks(mat, interval = 0.1)
#' @param mat Matrix to visualize.
#' @param interval Density of color breakpoints.
#' @return A vector listing the color breakpoints.
#' @author Leo Lahti, Olli-Pekka Huovilainen and Antonio Gusmao. Maintainer:
#' Leo Lahti <leo.lahti@@iki.fi>
#' @references L. Lahti et al.: Global modeling of transcriptional responses in
#' interaction networks. Submitted.
#' @keywords utilities
#' @export
#' @examples
#'  set.breaks(array(rnorm(100), dim = c(10, 10)), interval = .1) 
set.breaks <- function(mat, interval = 0.1) {
    if (max(abs(mat)) > 1) {
        m <- floor(max(abs(mat)))
    } else {
        m <- round(max(abs(mat)), nchar(1/interval) - 1)
    }
    
    mm <- m + interval/2
    vals <- seq(interval/2, mm, interval)
    # Note: the first and last values mimic infinity
    mybreaks <- c(-(m + 1e+06), c(-rev(vals), vals), m + 1e+06)
    mybreaks
}



ellipse <- function(centroid, covmat, confidence = 0.95, npoints = 100) {
    
    # Centroid: x0, y0 axes: a, b
    x0 <- centroid[[1]]
    y0 <- centroid[[2]]
    
    # Pick axis-wise stds
    a <- sqrt(diag(covmat))[[1]]
    b <- sqrt(diag(covmat))[[2]]
    
    theta <- seq(0, 2 * pi, length = npoints)
    
    # Confidence intervals (df=1: 1.96; df=2: 2.45)
    cint <- sqrt(qchisq(confidence, 2))
    
    # Determine point on the ellipse
    x <- x0 + cint * a * cos(theta)
    y <- y0 + cint * b * sin(theta)
    
    # Rotate alpha degrees from the x-axis: x <- x0 + a * cos(theta) * cos(alpha) - b
    # * sin(theta) * sin(alpha) y <- y0 + a * cos(theta) * sin(alpha) + b *
    # sin(theta) * cos(alpha)
    
    # Output ellipse points
    cbind(x, y)
    
}