R/vMask.method6.R

In vMask: Detect Small Changes in Process Mean using CUSUM Control Chart by v-Mask

vMask.method6 <-
function( 
 data,    #Type of inputed data must be a matrix with dim=c(m,n)
 mu0 = mean(data),    #Target value for process mean, with defult  mean(data)
 mu1,    #The althernative hypothesis H1: mu = mu1 (mu1=mu0+k*sigma) 
 sigma = sd(data),    #Supposed fixed (i.e. the process variation is assumed under control) and its defult is  sd(data)
 h = 5*sd(data),      #with defult  5*sigma  and it is "The length of decision making" and usully is in [4.5*sigma, 5*sigma].
 w = 2*sd(data),      #The ration of vertical unit into horisonal unit in CUSUM plot, with defult  2*sigma (this defoult couses theta is approximately 14 degrees).
 sleep = 1    #Sleep time (in secound)
 ){

  k = (mu1-mu0) / sigma
  d = h/k
  theta = atan( k/w )
  theta = rad2deg(theta)

  result <- vMask.method4(data, mu0, k, h, w, sleep)

  return( list( 
               d = d, 
               k = k, 
               theta = theta, 
               c = result$c, 
               OutControl = result$OutControl,  #Out-Control subsamples
               InControl = result$InControl    #In-Control subsamples
        )  
   )

}