uros-stegic / exptest2

exptest2 is R package that provides implementations of new (integral- and Kolmogorov- type) exponentiality tests introduced in paper [needs citation here] and also reimplementations of those provided by the package exptest.

exptest2

exptest2 is R package that provides implementations of new (integral- and Kolmogorov- type) exponentiality tests introduced in paper ^{[needs citation here]} and also reimplementations of those provided by the package exptest.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

Prerequisites

In ordrer to run and develop this software, you'll need the following things installed on your computer:

• R
• RStudio (this tutorial uses RStudio, console-type usage is not covered here)
• rcpp (available in CRAN repository)
• boost
• C++ compiler (with C++11 support)
• rbenchmark (optional, used to measure running times, also available in CRAN)

Installing

Building from source

To build this package, open it in RStudio and press the shortcut Ctrl+Shift+B which will compile the project and open new environment with this library included and ready to use.

Using precompiled binaries

First, you need to download the binary package here. After that, open RStudio and from the menu bar select Tools -> Install Packages... and in the popup window, in the Install From: dropdown menu select Package Archive File (.tar.gz). Then find and select the binary package you downloaded and hit Install.

Usage

ExpTest exports two functions for testing wether the given distribution comes from exponential family. See this example:

# Shows usage of both exp_test_integral and exp_test_kolmogorov functions
> library(ExpTest)
> x <- rexp(50)
> exp_test_integral(x)
[1] 0.003244134
> exp_test_kolmogorov(x)
[1] 0.07005984

For speed testing, you can use rbenchmark library that can be installed from CRAN. To measure speed of these functions, simply call the benchmark function like this:

# Measure running times of these two functions
> library(ExpTest)
> library(rbenchmark)
> x <- rexp(30)
> benchmark(replications = rep(1, 1, 1), exp_test_integral(x), exp_test_kolmogorov(x))
                    test replications elapsed relative user.self sys.self user.child sys.child
1   exp_test_integral(x)            1   0.018        1     0.019        0          0         0
2 exp_test_kolmogorov(x)            1   0.018        1     0.018        0          0         0

Authors

• Uros Stegic - Initial work - uros-stegic

License

This project is licensed under the GNU GPL-3 License - see the LICENSE file for details