Linguagem R aplicada ao planejamento fatorial em química

Abstract

The present work aimed to adapt the fabi_efeito and regression2 scripts from Octave/Matlab to the R language and apply them in the context of factorial design in chemistry using some examples, as well as using existing packages available for R in the optimization of an experiment and in the composition of a mixture. Factorial design applied to chemistry is used to help understand systems and optimize processes. For the calculations used in these methods, virtual spreadsheets or softwares capable of performing it are commonly used, with the free software RStudio being an alternative in the processing, analysis and visualization of these data, with a vast library of packages that can be applied to factorial design in chemistry. In this work, through the PacoteEfeitos package, an adaptation of the fabi_efeito and regression2 scripts, it was possible to calculate the percentage and probability of effects, obtain regression models, as well as the plot of graphs to visualize them and their parameters, in addition, through packages available for R, such as NlcOptim, rsm and mixexp, it was possible to obtain contour plots and response surfaces for a Doehlert-type design and contour plot for a mixture design, as well as their optimal regions and predicted responses.