R/curve_diff.R

In csgillespie/voltagefit: Voltage Curve Fitting

curve_diff = function(x, type, direction="Forward") {
  
  m1 = x[x$type == type[1] & x$direction == direction, ]
  m2 = x[x$type == type[2] & x$direction == direction, ]
  if(is.null(x$sim)) {
    data.frame(VG = m1$VG, diff = log10(m1$ID) - log10(m2$ID))  
    
  } else {
    message("sim")
    m1 = m1[order(m1$sim),]
    m2 = m2[order(m2$sim),]
    data.frame(VG = m1$VG, diff = log10(m1$ID) - log10(m2$ID), sim = m1$sim)
  }
}


#' Curve differences
#' 
#' Generate difference between two treatment/week effects. 
#' @param x    Data.frame as output by \code{\link{fit_manova}}.
#' @param type A vector of length two containing the comparision
#' @param direction Forward or Backward. 
#' @export
curve_diff_mean = function(x, type, direction="Forward") {
  means = mean(x)  
  curve_diff(means, type, direction)
}

#' @param n a positive integer; the number of samples to generate.
#' @rdname curve_diff_mean
#' @export
curve_diff_sample = function(x, type, n, direction="Forward") {
  unders = sample(x,  n=n)$samples
  curve_diff(unders, type, direction)
}