qualityTools (version 1.53)

paretoPlot: paretoPlot

Description

Display standardized effects and interactions of a 'facDesign' object in a pareto plot.

Usage

paretoPlot(fdo, threeWay = FALSE, abs = TRUE, decreasing = TRUE, na.last = NA,
           alpha = 0.05, response = NULL, xlim, ylim, xlab, ylab, main,
           single = TRUE, ...)

Arguments

fdo

an object of class facDesign

threeWay

logical. If TRUE, threeway-interactions are displayed as well.

abs

logical. If TRUE, absolute effects and interactions are displayed.

alpha

the significance level used to calculate the critical value

response

response variable. If the response data frame of fdo consists of more then one responses, this variable can be used to choose just one column of the response data frame. response needs to be an object of class character with length of

decreasing

logical. If TRUE, effects and interactions are sorted decreasing.

na.last

na.last

xlab

graphical parameter

ylab

graphical parameter

xlim

graphical parameter

ylim

graphical parameter

main

graphical parameter

single

a logical value.If TRUE a new graphic device will be opened for each column of the respond dataframe of fdo (response(fdo)). If set to FALSE par(mfrow) will be set internally. By default

...

graphical parameters

Value

• a list of effects for each response in the 'facDesign' object

Details

paretoPlot displays a pareto plot of effects and interactions for an object of class facDesign (i.e. 2^k full or 2^k-p fractional factorial design). For a given significance level alpha, a critical value is calculated and added to the plot. Standardization is achieved by dividing estimates with their standard error. For unreplicated fractional factorial designs a Lenth Plot is generated.

References

Design and Analysis of experiments - Volume2 - Advanced Experimental Design - Hinkelmann/Kempthorne

See Also

factors, fracDesign, facDesign

Examples

#factorial design with replications
#NA in response column and 2 replicates per factor combination
vp = fracDesign(k = 3, replicates = 2)  
#generate some data
y1 = 4*vp[,1] -7*vp[,2] + 2*vp[,2]*vp[,1] + 0.2*vp[,3] + rnorm(16)
y2 = 9*vp[,1] -2*vp[,2] + 5*vp[,2]*vp[,1] + 0.5*vp[,3] + rnorm(16)               
response(vp) = data.frame(y1,y2)
#show effects and interactions (nothing significant expected)
paretoPlot(vp)                         

#fractional factorial design --> Lenth Plot
vp = fracDesign(k = 4, gen = "D = ABC")
#generate some data
y = rnorm(8)                
response(vp) = y
#show effects and interactions (nothing significant expected)
paretoPlot(vp)