使用Agricolae包装的R中LSD没有给出值

Question

Hi: I am trying to calculate the LSD for some data in using the agricolae package in R. It seems straightforward enough from the documentation, but a part of the LSD is missing from the results.

I found that some others have had this problem, but was not able to find an answer to their questions. https://stat.ethz.ch/pipermail/r-help/2013-December/364391.html

Here is the code I am using, and some data I generated which replicates the results I am getting on my real data.

library(agricolae)

## Create Data Frame 

factor <- c(1,1,1,1,2,2,2,2,3,3,3)
var <- c(22,23,24,27,13,17,18,10,31,29,33)

df <- data.frame(factor,var)

## Run LSD test 

model <- aov(var ~ factor, df)
out <- LSD.test(model, "factor")

## Results 

out$statistics$LSD
NULL

Answer 1

No LSD value is given in the output because it appears that this output is only included when you have perfectly balanced data, and the data you have provided are unbalanced (ie 4 group "1" replications, 4 group "2" replications, but only 3 group "3" replications). When the data are balanced, the LSD value is given, for example:

library(agricolae)

## Create Data Frame 

factor <- c(1,1,1,1,2,2,2,2,3,3,3,3)
var <- c(22,23,24,27,13,17,18,10,31,29,33,27)

df <- data.frame(factor,var)

## Run LSD test 

model <- aov(var ~ factor, df)
out <- LSD.test(model, "factor")

## Results 

out$statistics$LSD
11.04747

Answer 2

Try to make the factor vector as factor after you make the data frame and before build the model. Here is the full code:

library(agricolae)
factor <- c(1,1,1,1,2,2,2,2,3,3,3)
var <- c(22,23,24,27,13,17,18,10,31,29,33)
df <- data.frame(factor, var) 
df$factor<-as.factor(df$factor) #make the factor "factor"
model<-aov(var~factor, data=df)
(LSD.test(model, "factor"))$groups