vMask.method4: Method 4 for V-Mask Implementation: Using mu0, k, h and w

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

Description

The V-mask method in vMask Package, is considered on the basis of variety of situations/information by different functions. One of these functions is vMask.method1 which can plot the V-mask based on μ_0, k and h and w.

Usage

vMask.method4(data, mu0 = mean(data), k = 0.5 * sd(data), h = 5 * sd(data), 
w = 2 * sd(data), sleep = 1)

Arguments

data	Type of inputted data must be a matrix with dim=c(m,n), in which m is size of samples and n is size of each subsample. Meanwhile, data can be a numerical vector where its elements are sub-sample means.
mu0	The target value for process mean, with default mean(data).
k	With default 0.5 σ and it depends on the form of the alternative hypothesis H_1: μ = μ_1 \ (μ_1=μ_0+k σ).
h	The length of decision making (h) is equal to the vertical distance between the origin and the upper (or lower) arm of V-mask. Its default is h = 5 sd(data), since usually h \in [4.5 σ, 5 σ] is proposed.
w	The ration of vertical unit into horizontal unit in CUSUM plot, with default w = 2 \times sd(data) (this default causes θ to be approximately equal to 14 degrees).
sleep	Sleeping time (in second) of the program between showing figures. This time needs for see the result of checking i-th point on CUSUM control chart by V-mask. Also, if sleep="PressEnter", then the user must press key [Enter] in 'R Console' window to continue/check the next cumulative summation point on chart with V-mask.

Value

d	The distance between two following points: (1) O, i.e. the latest cumulative sum point (which duty of the V-mask is checking this point), and (2) P, i.e. the junction points of V-mask arms.
theta	half of the angle formed by the V-mask arms (in degrees).
h	the vertical distance between the origin and the upper (or lower) arm of V-mask.
c	A vector of the cumulative differences between statistic (mean) values and the mean target value; i.e. c=(c_1, ..., c_m) where c_i=∑_{j=1}^{i} (\bar{x}_j - μ_0) .
OutControl	The number of Out-of-control points in CUSUM chart.
InControl	The number of In-control points in CUSUM chart.

See Also

vMask.method3

Examples

m = 35	#The size of samples
n = 2 	#The size of each subsample
set.seed(1234)
Data = matrix(rnorm(m*n, 0,1), nrow=m)
head(Data)

vMask.method4( data=Data, s=0 )
vMask.method4( data=Data, k=.3, h=2.5, w=1, s=0 )
vMask.method4( data=Data, k=.3, h=2.5, w=2, s=0 ) #Compare with the previous line

vMask documentation built on May 1, 2019, 10:21 p.m.