CRAN status Last-changedate codecov Build Status Lifecycle: stable

midrangeMCP

midrangeMCP is an “R” package designed to present four new multiple comparison tests. This work was developed during the doctorate in Statistics and Agricultural Experimentation, by the Federal University of Lavras (UFLA/BRAZIL). The supervisor of this work was Professor Daniel Furtado Ferreira. I am currently as Professor of Statistics at the Federal University of São João del-Rei (UFSJ/BRAZIL). Today, we have the collaboration of a student of scientific initiation, Diego Arthur Bispo Justino de Oliveira, student of Mechatronics Engineering at UFSJ.

Four tests are presented, three based on the distribution of the externally studentized midrange and one based on the distribution of the externally studented range, well documented distribution in the literature. There are many distributions based on this last distribution, the difference for our case is that it was used for a test based on grouping methods for multiple comparisons, something similar to what was done with the Scott-Knot test.

The tests based on the distribution of the externally studentized midrange are the Tukey Midrange test (TM test), the SNK Midrange test (SNKM test) and the Means Grouping based on Midrange test (MGM test). Finally, the test based on the distribution of the externally studentized range which is the Means Grouping based on the Range test (MGR test).

Finally, a versatility of the package is the use of the GUI. This feature provides users unfamiliar with the R language when using the midrangeMCP package.

These tests are available in the thesis of the work, and their performance evaluations are presented. The articles of these are in the submission stage, one of which has already been approved by the Journal of Science and Agrotechnology.

Instalation

To install the midrangeMCP package via CRAN:

install.packages("midrangeMCP")

To install via GitHub:

install.packages("devtools")
install_github("bendeivide/midrangeMCP")

Functions

The package’s functions are: MRbarplot, MRwrite, guimidrangeMCP and MRtest. The latter is the main function. From the arguments inserted in it, the other functions can be performed. In the case of the guimidrangeMCP function, once executed, it will open the GUI. From there the database and the necessary arguments for the functions can be entered.

Citation

The package article is still being submitted. But the works mentioned above can assist in citing the package. Still, as the midrangeMCP package is available on CRAN, quote it this way:

#> Warning in citation("midrangeMCP"): no date field in DESCRIPTION file of package
#> 'midrangeMCP'
#> Warning in citation("midrangeMCP"): could not determine year for 'midrangeMCP'
#> from package DESCRIPTION file
#> 
#> To cite package 'midrangeMCP' in publications use:
#> 
#>   Ben Deivide and Daniel Furtado (NA). midrangeMCP: Multiples
#>   Comparisons Procedures Based on Studentized Midrange and Range
#>   Distributions. https://bendeivide.github.io/midrangeMCP/,
#>   https://github.com/bendeivide/midrangeMCP.
#> 
#> A BibTeX entry for LaTeX users is
#> 
#>   @Manual{,
#>     title = {midrangeMCP: Multiples Comparisons Procedures Based on Studentized Midrange and Range Distributions},
#>     author = {Ben Deivide and Daniel Furtado},
#>     note = {https://bendeivide.github.io/midrangeMCP/, https://github.com/bendeivide/midrangeMCP},
#>   }

Website

For a complete description of the midrangeMCP package, visit our website at bendeivide.github.io/midrangeMCP/

News