Fourth Corner Algorithm in R - r

Question

This is a question about the fourthcorner algorithm in R. It's designed to measure the relationship between three different tables: an n x m table (table R) of m environmental variables (columns) at n sites (rows), an n x p table (table L) of p abundances (columns) at n sites (rows), and a p x s table (table Q) of s traits (columns) for p species (rows).
The fourthcorner function is in the package ade4.
All three of my dataframes are binary (0s and 1s denoting the presence or absence of a variable, a species at a site, or a trait, respectively). I've tried using "yes" and "no" instead of 0s and 1s without success.
Here are some example matrices in the format I'm using:
tabQ
Trait1 Trait2 Trait3 Trait4
Sp1 0 1 0 0
Sp2 0 1 0 0
Sp3 1 0 1 0
Sp4 1 0 1 0
Sp5 0 1 0 0
Sp6 0 1 0 0
Sp7 0 0 0 1
Sp8 0 0 0 1
tabR
EnV1 EnV2 EnV3 EnV4
Site1 1 1 1 1
Site2 1 1 0 1
Site3 0 1 0 1
Site4 1 1 1 1
Site5 1 1 0 1
Site6 0 1 0 0
Site7 0 1 0 1
Site8 0 1 0 1
Site9 1 1 1 1
Site10 1 1 0 1
Site11 1 1 1 1
Site12 0 1 0 0
Site13 1 1 0 1
Site14 1 1 0 1
Site15 0 1 0 1
Site16 1 1 0 1
Site17 0 1 0 1
Site18 1 1 1 1
Site19 1 1 0 1
Site20 1 1 0 1
tabL
Sp1 Sp2 Sp3 Sp4 Sp5 Sp6 Sp7 Sp8
Site1 1 1 0 0 0 0 0 0
Site2 1 1 0 0 0 0 0 0
Site3 1 1 0 0 0 0 0 0
Site4 1 0 0 0 0 0 0 1
Site5 1 1 0 0 0 0 0 0
Site6 1 0 0 0 1 0 0 0
Site7 1 0 0 0 0 0 0 0
Site8 0 0 0 0 1 0 0 0
Site9 1 0 0 0 0 0 0 0
Site10 1 1 0 0 0 0 0 0
Site11 0 0 1 1 0 0 0 0
Site12 0 0 0 0 0 1 0 0
Site13 1 0 0 0 0 0 0 0
Site14 0 0 0 0 1 0 0 0
Site15 1 1 0 0 0 0 0 0
Site16 1 1 0 0 0 0 0 0
Site17 1 0 0 0 0 0 0 0
Site18 0 0 1 0 0 0 0 0
Site19 1 0 0 0 0 0 0 0
Site20 1 1 0 0 0 0 1 0
I read these dataframes into R from text files, and I specify that the first column is row names.
This is the error I get when I try to use the fourthcorner function on my matrices:
fourth1=fourthcorner(tabR,tabL,tabQ,nrepet=1)
Error in apply(sim, 2, function(x) length(na.omit(x))) :
dim(X) must have a positive length
I don't understand where the problem lies, is it a formatting issue? If so, should I reformat one of the matrices? Which one is causing the trouble? Or can I not use binary traits and environmental variables for this function? In other words, can I solve this problem by changing a piece of code, or is it impossible to use this function for this question?
As an additional tidbit of information, I did email the author of the function, but unfortunately I did not understand his response fully, possibly because my R skills still leave much to be desired. Here is his response if it is helpful:
Q could contain quantitative or qualitative traits. In R, qualitative traits should be coded as factors to obtain adapted statistics (i.e. chi2 or eta2). If you code qualitative variables as dummy variables, they would be considered as quantitative.
Thank you very much to any and all insight.

Answer 1

I noted that your example fails only nrepet is equal to one, so if you can use any other positive number you should be fine.
However, if you do need nrepet=1, you should contact with the author of ade4 and ask to him/her to fix the fourthcorner function code. I traced back the error and found that fourthcorner calls as.krandtest with sim = res$tabD[-1,] where res$tabD is a matrix with nrepet+1 rows. When nrepet=1 and you remove one row from a two-row matrix, R automatically converts the resulting one-row matrix into a vector, but as.krandtest function expects sim to be a matrix and thus raises the error.
Here is your input data just in case somebody else would like to answer your question:
tabR
structure(list(EnV1 = c(1L, 1L, 0L, 1L, 1L, 0L, 0L, 0L, 1L, 1L,
1L, 0L, 1L, 1L, 0L, 1L, 0L, 1L, 1L, 1L), EnV2 = c(1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L), EnV3 = c(1L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 1L, 0L, 1L, 0L,
0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L), EnV4 = c(1L, 1L, 1L, 1L, 1L,
0L, 1L, 1L, 1L, 1L, 1L, 0L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L)), .Names = c("EnV1",
"EnV2", "EnV3", "EnV4"), row.names = c("Site1", "Site2", "Site3",
"Site4", "Site5", "Site6", "Site7", "Site8", "Site9", "Site10",
"Site11", "Site12", "Site13", "Site14", "Site15", "Site16", "Site17",
"Site18", "Site19", "Site20"), class = "data.frame")
tabL
structure(list(Sp1 = c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 0L, 1L, 1L,
0L, 0L, 1L, 0L, 1L, 1L, 1L, 0L, 1L, 1L), Sp2 = c(1L, 1L, 1L,
0L, 1L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 1L, 1L, 0L, 0L, 0L,
1L), Sp3 = c(0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L,
0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L), Sp4 = c(0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L),
Sp5 = c(0L, 0L, 0L, 0L, 0L, 1L, 0L, 1L, 0L, 0L, 0L, 0L, 0L,
1L, 0L, 0L, 0L, 0L, 0L, 0L), Sp6 = c(0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L
), Sp7 = c(0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L), Sp8 = c(0L, 0L, 0L, 1L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
0L)), .Names = c("Sp1", "Sp2", "Sp3", "Sp4", "Sp5", "Sp6",
"Sp7", "Sp8"), row.names = c("Site1", "Site2", "Site3", "Site4",
"Site5", "Site6", "Site7", "Site8", "Site9", "Site10", "Site11",
"Site12", "Site13", "Site14", "Site15", "Site16", "Site17", "Site18",
"Site19", "Site20"), class = "data.frame")
tabQ
structure(list(Trait1 = c(0L, 0L, 1L, 1L, 0L, 0L, 0L, 0L), Trait2 = c(1L,
1L, 0L, 0L, 1L, 1L, 0L, 0L), Trait3 = c(0L, 0L, 1L, 1L, 0L, 0L,
0L, 0L), Trait4 = c(0L, 0L, 0L, 0L, 0L, 0L, 1L, 1L)), .Names = c("Trait1",
"Trait2", "Trait3", "Trait4"), row.names = c("Sp1", "Sp2", "Sp3",
"Sp4", "Sp5", "Sp6", "Sp7", "Sp8"), class = "data.frame")