outliers.ils

Procedure for detecting funcitonal outliers.

It performs interlaboratory studies (ILS) to detect those
laboratories that provide non-consistent results when comparing to others. It
permits to work simultaneously with various testing materials, from standard
univariate, and functional data analysis (FDA) perspectives. The univariate
approach based on ASTM E691-08 consist of estimating the Mandel's h and
k statistics to identify those laboratories that provide more significant
different results, testing also the presence of outliers by Cochran and Grubbs
tests, Analysis of variance (ANOVA) techniques are provided (F and Tuckey
tests) to test differences in means corresponding to different laboratories per
each material. Taking into account the functional nature of data retrieved in
analytical chemistry, applied physics and engineering (spectra, thermograms,
etc.). ILS package provides a FDA approach for finding the Mandel's k and h
statistics distribution by smoothing bootstrap resampling.

Miguel Flores

Interlaboratory Study

Miguel Flores 

Ruben Fernandez 

Salvador Naya 

Javier Tarrio-Saavedra 

outliers.ils function

<dl><dt>x</dt>
<dd>An object of classs <code>fdata</code>.</dd>
<dt>fdep</dt>
<dd>Type of depth measure, by default depth.mode.</dd>
<dt>trim</dt>
<dd>The percentaje of the trimming, by default is 1%.</dd></dl>

Arguments

Detecting outliers for functional dataset — outliers.ils

<dl>

<dt>x</dt>
<dd>An object of classs <code>fdata</code>.</dd>


<dt>fdep</dt>
<dd>Type of depth measure, by default depth.mode.</dd>


<dt>trim</dt>
<dd>The percentaje of the trimming, by default is 1%.</dd>

</dl>

outliers.ils: Detecting outliers for functional dataset

Description

Usage

Arguments