Creation of an R package for multivariate control graphics with variable dimention

Abstract

Multivariate Control charts are a valuable contribution in the process control field, the control chart  Hotelling is the most commonly used in this field. The present paper is based on the methodology proposed by the authors Aparisi, F., Epprecht E. K., Ruiz-Barzola, O., about Variable Dimension and Double Dimension Control Chart . The main contribution of the article is the development of the package in R based on the graduation project of Barahona-Alvarado Jhonny. The package in R perform multivariate control charts and calculates their respective ARL, using the graphics  Hotelling, Doble Dimension  and Variable Dimension these last two are variations made the traditional  that get better performance and at the same time reduce the cost of sampling. For validation of the results the simulación was performed of p=p1+p2 multivariate normal variables, p1 variables easy to measure or cheap and p2 variables difficult to measure or expensive. It was obtained as a result the package in R, which consists of several functions that performs the three control charts. It was further confirmed that variable dimension graph overtakes performance of traditional control graph  with the p variables, the dimension graph overtakes performance graphic  with p1 variables and sometimes overtakes the traditional  with the p variables. The control graphs   and reduce the cost of sampling.