R Package DoE.base for Factorial Experiments

Abstract

The R package DoE.base can be used for creating full factorial designs and general factorial experiments based on orthogonal arrays. Besides design creation, some analysis functionality is also available, particularly (augmented) half-normal effects plots. In addition to this specific functionality, the package provides convenience features for analyzing experimental designs and the infrastructure for a suite of further packages on designing and analyzing experiments. This infrastructure is available for use also by further design of experiments packages.

Structured plant metabolomics for the simultaneous exploration of multiple factors

“…The experimental design was based on an orthogonal array with 96 runs created with the free open-source R package DoE.base47 as described in the supplementary material. Given the size of this experiment, tests for effects of 2-level factors at significance level 5% can detect effects of size “one standard deviation” with about 99% power, effects of size “half a standard deviation” with about 68% power, and effects of size “0.75 standard deviations” with about 95% power.…”

DoE.MIParray: An R Package for Algorithmic Creation of Orthogonal Arrays

“…This brief section comments on a small selection of those packages that could be used for similar applications as DoE.MIParray: relatively small experiments with several factors, some or all of which are qualitative, and all of which have a relatively small number of levels, not necessarily the same for different factors. DoE.base (see [4]) handles such situations using catalogued orthogonal arrays, possibly optimizing balance by column selection from these. DoE.MIParray algorithmically creates wellbalanced arrays by mixed integer optimization; the author is not aware of any other software tool, in R or elsewhere, that implements the same approach.…”

“…(For experiments with many factors, explicit replication may be replaced by implicit replication, see e.g. [4]). The above small experiment only has the m = 3 factors bp (baking powder brand) with 3 levels, oven with 2 levels and recipe with 2 levels.…”

“…A full factorial experiment would have needed 576 runs, which was considered infeasible. The design plan used in [11] was obtained using column selection optimization based on a catalogued 72 run plan of R package DoE.base (see [4]). Suitable use of R package DoE.MIParray creates a more balanced plan; thus, had DoE.MIParray already been available when conducting the experiment reported in [11], a more balanced experiment could have been run.…”

Determinants of the Price of Housing in the Province of Alicante (Spain): Analysis Using Quantile Regression

“…To estimate the relative importance of each variable in the regression model, a method developed by Lindeman et al [91] was used, which allows for the breakdown of the explanatory variability of each independent variable. The averaged over orderings method (lmg metric) included in the "relaimpo" package (version 2.2-3) of R [92], developed by Grömping [93], was used. One To estimate the relative importance of each variable in the regression model, a method developed by Lindeman et al [91] was used, which allows for the breakdown of the explanatory variability of each independent variable.…”