This function allows us to compare the treatments averages or the adding of their ranges with the minimal significant difference which can vary from one comparison to another one.
order.group( trt, means, N, MSerror, Tprob, std.err, parameter = 1, snk = 0, DFerror = NULL, alpha = NULL, sdtdif = NULL, vartau = NULL, console )
| trt | Treatments |
|---|---|
| means | Means of treatment |
| N | Replications |
| MSerror | Mean square error |
| Tprob | minimum value for the comparison |
| std.err | standard error |
| parameter | Constante 1 (Sd), 0.5 (Sx) |
| snk | Constante = 1 (Student Newman Keuls) |
| DFerror | Degrees of freedom of the experimental error |
| alpha | Level of risk for the test |
| sdtdif | standard deviation of difference in BIB |
| vartau | matrix var-cov in PBIB |
| console | logical, print output |
The output is data frame.
Treatment Levels, Factor
height, Numeric
groups levels, Factor
replications, Numeric
Standard error, Numeric
This function was changed by orderPvalue function that use agricolae. Now the grouping in agricolae is with the probability of the treatments differences and alpha level.
It is considered 81 labels as maximum for the formation of groups, greater number will not have label.
library(agricolae) treatments <- c("A","B","C","D","E","F") means<-c(20,40,35,72,49,58) std.err<-c(1.2, 2, 1.5, 2.4, 1, 3.1) replications <- c(4,4,3,4,3,3) MSerror <- 55.8 value.t <- 2.1314 groups<-order.group(treatments,means,replications,MSerror,value.t,std.err,console=FALSE) print(groups)#> trt means M N std.err #> 1 D 72 a 4 2.4 #> 2 F 58 b 3 3.1 #> 3 E 49 bc 3 1.0 #> 4 B 40 cd 4 2.0 #> 5 C 35 d 3 1.5 #> 6 A 20 e 4 1.2