Statistical Procedures for Agricultural Research

Original idea was presented in the thesis "A statistical analysis tool for agricultural research" to obtain the degree of Master on science, National Engineering University (UNI), Lima-Peru. Some experimental data for the examples come from the CIP and others research. Agricolae offers extensive functionality on experimental design especially for agricultural and plant breeding experiments, which can also be useful for other purposes. It supports planning of lattice, Alpha, Cyclic, Complete Block, Latin Square, Graeco-Latin Squares, augmented block, factorial, split and strip plot designs. There are also various analysis facilities for experimental data, e.g. treatment comparison procedures and several non-parametric tests comparison, biodiversity indexes and consensus cluster.


Changes in new version

Version agricolae 1-2.9 (January 4, 2019) function include argument label size with cex=NULL HSD.test include parameter unbalanced, equal TRUE is not equal replication

Version agricolae 1-2.8 (September 12, 2017)

The function is again in agricolae. It is equivalent to the orderPvalue function in functional terms.

Version agricolae 1-2.7 (August 30, 2017)

In the post.hoc tests, the grouping of treatments are formed according to the probability of the difference between treatments and the alpha level.

The affected functions were BIB.test, DAU.test, duncan.test, durbin.test, friedman, HSD.test, kruskal, LSD.test, Median.test, PBIB.test, REGW.test, scheffe.test, SNK. Test, waller.test and waerden.test. Now there is good correspondence between the grouping and the pvalue.

A new function ( is included in agricolae for the graphs of treatment groups and their variation by range, interquartil range, Standard deviation and standard error.

The RANN package of suggestions was removed.

Updated documentation.

Version agricolae 1-2.6 (August 4, 2017)

Documentation check.

Version agricolae 1-2.5 (July 20, 2017)

  1. Add model object in output PBIB.test function.
  2. procedure duncan.test is better, the limitations in convergence were corrected.
  3. The influence in AMMI (type=3) is relative neighbor graph as a sub-graph.
  4. The post hoc nonparametrics tests (kruskal, friedman, durbin and waerden) are using the criterium Fisher's least significant difference (LSD)

Version agricolae 1-2.4 (June 12, 2016)

  1. Add suggests packages: RANN and rgeos to plot AMMI
  2. Concordance index in correlation function(), additional arg (method="lin").
  3. New function orderPvalue(). Grouping the treatments in a comparison with p.value minimum value (alpha)
  4. Test LSD.test and kruskal the adjust P.value (holm, hommel, hochberg, bonferroni, BH, BY, fdr). The comparison in pairs and groups give similar results.

Version agricolae 1-2.3 (October 6, 2015)

  1. REGW.test(). New function for multiple comparisons of treatments. (Ryan, Einot and Gabriel and Welsch)
  2. diffograph(). New function: Mean-mean scatter plot, test: Bonferroni, Fisher, Duncan, Student-Newman-Keul Tukey, Kruskal-Wallis, Friedman and Waerden test.
  3. Changes in all comparison means, add parameters to facility function diffograph.
  4. Added randomization parameter (TRUE or FALSE) in all design function.
  5. Update Tutorial

Version agricolae 1-2.2 (August 12, 2015)

  1. Now in the frequency table shows the relative frequency as a percentage, the function is table.freq or summary( graph.freq or hist object)
  2. The histogram class is added to graph.freq and it can use the package HistogramTools
  3. The function design.bib create optimal design, use function optBlock(algDesign)
  4. sketch option in design: rbcd, lsd, graeco, youden, bib

Version agricolae 1-2.1 (August 25, 2014)

  1. Move packages from Suggests to Imports
  2. AUDPS. The Area Under the Disease Progress Stairs.
  3. AMMI stability value (ASV) and Yield stability index (YSI)
  4. Design youden
  5. Now the PBIB.test function uses missing values.

Version agricolae 1-2.0 (June 30, 2014)

  1. AMMI: aditional parameters PC=FALSE or TRUE, output principal components, check error equal cero.
  2. plot.AMMI: graphic aditional parameters lwd = 1.8, length = 0.1 to arrow function
  3. simulation.model: aditional parameter console=FALSE or TRUE, output in console
  4. resampling.model: aditional parameter console=FALSE or TRUE, output in console
  5. stability.par: aditional parameter console=FALSE or TRUE, output in console
  6. stability.nonpar: aditional parameter console=FALSE or TRUE, output in console

Version agricolae 1-1.9 (June 17, 2014)

  1. PBIB new parameter: group=TRUE PBIB.test(block,trt,replication,y,k, method=c("REML","ML","VC"), test = c("lsd","tukey"), alpha=0.05, console=FALSE, group=TRUE) when you have many treatments to use group=FALSE.
  2. design.rcbd(..., continue=FALSE) continue=TRUE or FALSE, continuous numbering of plot.
  3. Median.test. New function for multiple comparisons of treatments with Median.
  4. Now, AMMI function checks the minimum number of environments and genotypes. Now use console=TRUE or FALSE to output in screen. the graphs are produced by the plot function.
  5. plot.AMMI() or plot() functions generate plot of the AMMI with others principal components. type=1 (biplot), type=2 (triplot) and type=3 (influence genotype)
  6. Changed parameters by default "first = TRUE" in designs: rcbd, ab, split and lsd.
  7. Now vignettes in agricolae.
  8. change name ogive.freq by ojiva.freq, the parameters are same.
  9. AUDPC the evaluation parameter now can be numeric vector. To see help(audpc)

Version agricolae 1-1.8 (February 21, 2014)

  • zigzag(outdesign) The new function applied to designs: rcbd, lsd, graeco, split, strip, ab, alpha, bib, cyclic, lattice, dau. The outdesign is the output book the function design.###(). The function zigzag change the order number plots in serpentine form.

Randomized complete design.

trt<-LETTERS[1:5] outdesign<-design.rcbd(trt,r=4, serie=2) book<-outdesign$book

t(matrix(book[,1],5)) [,1] [,2] [,3] [,4] [,5] [1,] 101 102 103 104 105 [2,] 201 202 203 204 205 [3,] 301 302 303 304 305 [4,] 401 402 403 404 405 fieldbook <- zigzag(outdesign) t(matrix(fieldbook[,1],5)) [,1] [,2] [,3] [,4] [,5] [1,] 101 102 103 104 105 [2,] 205 204 203 202 201 [3,] 301 302 303 304 305 [4,] 405 404 403 402 401

  • Now, all design functions have two output objects: parameters and book, the parameters contain initial values that will allow reproduce the design and book contain field book.
  • The alpha and lattice designs have additionally two objects: statistics and field sketches.
  • BIB have the statistics with parameters and field book.
  • Cyclic have the stetches with parameters and field book.

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.