powerplot

Plots powers for the Kruskall-Wallis test, via Monte Carlo and two approximations.

A collection of multivariate nonparametric methods, selected in part to support an MS level course in nonparametric statistical methods. Methods include adjustments for multiple comparisons, implementation of multivariate Mann-Whitney-Wilcoxon testing, inversion of these tests to produce a confidence region, some permutation tests for linear models, and some algorithms for calculating exact probabilities associated with one- and two- stage testing involving Mann-Whitney-Wilcoxon statistics. Supported by grant NSF DMS 1712839. See Kolassa and Seifu (2013) <doi:10.1016/j.acra.2013.03.006>.

John E Kolassa

MultNonParam

Multivariate Nonparametric Methods

John E. Kolassa

Stephane Jankowski

powerplot function

<dl><dt>numgrps</dt>
<dd>Number of groups to compare</dd>
<dt>thetadagger</dt>
<dd>Direction of effect</dd>
<dt>nnvec</dt>
<dd>vector of numbers per group.</dd>
<dt>nmc</dt>
<dd>Number of Monte Carlo trials</dd>
<dt>targetpower</dt>
<dd>Target power for test</dd>
<dt>level</dt>
<dd>level for test.</dd></dl>

Arguments

Power Plot — powerplot

<dl>

<dt>numgrps</dt>
<dd>Number of groups to compare</dd>


<dt>thetadagger</dt>
<dd>Direction of effect</dd>


<dt>nnvec</dt>
<dd>vector of numbers per group.</dd>


<dt>nmc</dt>
<dd>Number of Monte Carlo trials</dd>


<dt>targetpower</dt>
<dd>Target power for test</dd>


<dt>level</dt>
<dd>level for test.</dd>

</dl>

powerplot: Power Plot

Description

Usage

Arguments