How to fix "Error in x[i, ] : subscript out of bounds" for ANOVA, Tukey HSD

Question

I am receiving the error "Error in x[i, ] : subscript out of bounds while using the TukeyC library for R.

I am attempting to run an ANOVA followed by the Tukey HSD post-hoc test. The code (below) works fine for my initial dataset richORIG.

library(TukeyC)

avORIG <- with(richORIG, aov(rich ~ ClimDiv_ORIG, data=richORIG))
summary(avORIG)                             
tkORIG <- TukeyC(x=avORIG, which='ClimDiv_ORIG')
summary(tkORIG)
plot(tkORIG)

Resulting in:

avORIG <- with(richORIG, aov(rich ~ ClimDiv_ORIG, data=richORIG))

> summary(avORIG)
               Df Sum Sq Mean Sq F value Pr(>F) 
ClimDiv_ORIG    8  488.7   61.09   76.17 <2e-16 ***
Residuals    1413 1133.2    0.80                                    

> tkORIG <- TukeyC(x=avORIG, which='ClimDiv_ORIG')
trace: TukeyC(x = avORIG, which = "ClimDiv_ORIG")

> summary(tkORIG)
Goups of means at sig.level = 0.05 
   Means G1 G2 G3 G4 G5
12  5.10  a            
23  4.70     b         
11  4.68     b         
22  4.50     b         
13  4.24        c      
21  4.03        c      
33  3.26           d   
32  3.14           d   
31  2.38              e

> plot(tkORIG)  ##I can't post the picture of the plot w/o appropriate reputation

I try this again with my rich2080 dataset

av2080 <- with(rich2080, aov(rich ~ ClimDiv_2080, data=rich2080))
summary(av2080)
tk2080 <- TukeyC(x=av2080, which='ClimDiv_2080')
summary(tk2080)
plot(tk2080)

But get the error below

> av2080 <- with(rich2080, aov(richtimestep ~ ClimDiv_2080, data=rich2080))

> summary(av2080)
               Df Sum Sq Mean Sq F value   Pr(>F)
ClimDiv_2080    8   16.2  2.0264   7.574 5.97e-10
Residuals    1416  378.9  0.2676                 

> tk2080 <- TukeyC(x=av2080, which='ClimDiv_2080')
trace: TukeyC(x = av2080, which = "ClimDiv_2080")

Error in x[i, ] : subscript out of bounds

The code works for all of my other data-sets (ex: simpORIG, simp2080, shanORIG, shan2080 etc.)

I came across this StackOverflow Q/A in reference to my question, but I cannot figure out what code to change in my particular circumstance.

Here are the options available from the debug menu, but as I sift through the ls(), I don't know which value to change...

Enter a frame number, or 0 to exit   

1: TukeyC(x = av2080, which = "ClimDiv_2080")
2: TukeyC.aov(x = av2080, which = "ClimDiv_2080")
3: make.TukeyC.test(r = r, MSE = MSE, m.inf = m.inf, ord = ord, sig.level
4: make.TukeyC.groups(dif)

How would I go about fixing this error to get my results?

score 0 · Answer 1 · edited Jul 14 '16 at 12:52

I had this exact error, but with the TukeyHSD() function in the basic stats package. I have no idea why it happened - in short, yesterday it was working and today it's not. Really really strange...

I removed the stats package (BAD IDEA) with the hope of re-installing it, but couldn't. Eventually I just re-started by downloading the latest R version and starting again.

Sorry that's not more specific, but I couldn't find an explanation for why the function suddenly wasn't working, and I presume TukeyC() works in a very similar way to TukeyHSD()

It was like in Jurrasic Park when they had to switch the power back on: Long, complicated and scary...

How to fix "Error in x[i, ] : subscript out of bounds" for ANOVA, Tukey HSD

1 Answers1