# Overview

The purpose of PH1XBAR is to build three types of Phase I Shewhart control charts:

1. Phase I Shewhart X-bar control chart with a balanced one-way random effects model (doi:10.1002/qre.2793).

Phase I Shewhart individual control chart for the iid case (doi:10.1080/08982112.2021.1878220).

Phase I individual control chart with an ARMA model.

# Installation

## Install from CRAN

PH1XBAR is published on CRAN, so we recommend users installing it in a regular way as follows

`install.packages("PH1XBAR")`

## Install from GitHub

PH1XBAR is still under development, so if users are more interested in the experimental version, there is an alternative installation through Github as follows

```
install.packages("devtools")
devtools::install_github("bolus123/PH1XBAR")
```

Note that for Windows users, Rtools may need to be installed in advance. Please choose the right version of Rtools which is corresponding to your R and Rstudio. The detailed instruction is introduced: https://cran.r-project.org/bin/windows/Rtools/

For Mac and Linux users, please follow the instruction: https://www.r-project.org/nosvn/pandoc/devtools.html

## Install from local

Users can also download our release, PH1XBAR_x.y.z.tar.gz, from our homepage on CRAN or Github and then install it from your local path as follows

`install.packages('path_to_file/PH1XBAR_x.y.z.tar.gz', repos = NULL, type="source")`

# Usage

Before using any functions, PH1XBAR may need to be loaded into R

PH1XBAR provides a function to build Phase I X-bar chart with variance components model as follows

```
data(grinder_data)
PH1XBAR(grinder_data)
```

Notice that the variance estimator in the control chart must be S or MR. Also, PH1XBAR provides a function to get the corrected charting constant as follows

```
# S-based estimator involved
getCC.XBAR(FAP0 = 0.1, m = 30, var.est = 'S')
# MR-based estimator involved
getCC.XBAR(FAP0 = 0.1, m = 30, var.est = 'MR')
```

PH1XBAR provides a function to build Phase I individual chart with an ARMA model as follows

```
data(preston_data)
# using the default setting whose FAP0 = 0.1
PH1ARMA(preston_data)
# using known parameters with FAP0 = 0.1
PH1ARMA(preston_data, case = 'K')
```

Also, PH1XBAR provides a function to get the ARMA corrected charting constant as follows

```
# double simulation gets involved if estimators are unknown
getCC.ARMA(FAP0 = 0.1, double.sim = TRUE)
# single simulation gets involved if estimators are known
getCC.ARMA(FAP0 = 0.1, double.sim = FALSE)
```

More details are on the manual.