Skip to content

Help with RDA analysis, function ''varpart'' in vegan

3 messages · Sophi Marmen, Eric Berger, Bert Gunter

Sophi Marmen
# 
I'm trying to perform a RDA analysis in the vegan package using the
"varpart"  function.

I have a matrix of community structure data in different sites (rows),
which I want to explain using 3 matrices of environmental data. The 3
matrices are:

water quality parameters;

local land use variables;

total land use variables.

In each matrix, the number of the rows is the same (the sites) and there
are different column number (measured varibakes).

I use the follow script:

WQ_RDA <- read.csv("F:/Sher_sophi_new3/RDA/WQ_RDA_1.csv", row.names = 1)
LU2_loc_RDA <- read.csv("F:/Sher_sophi_new3/RDA/LU2_loc_RDA_1.csv",
row.names = 1)
LU2_bas_RDA <- read.csv("F:/Sher_sophi_new3/RDA/LU2_basin_RDA_1.csv",
row.names = 1)
comm <- read.csv("F:/Sher_sophi_new3/RDA/Final_true_OTUs_97_1.csv",
row.names = 1)

x1 = as.matrix(WQ_RDA)
x2 = as.matrix(LU2_loc_RDA)
x3 = as.matrix(LU2_bas_RDA)
comm = as.matrix(t(comm))

RDA_Ger = varpart(comm, x1, x2, x3, transfo="hellinger", scale = FALSE,
                  na.action = na.omit


I get this error message:

Error in as.vector(x, mode) :
  cannot coerce type 'closure' to vector of type 'any'

Can anyone help me figure out what am I doing wrong?

Thanks,
Sophi
Eric Berger
# 
I am not familiar with the vegan package, so I am just making a guess here.
If 'na.action=na.omit' is part of the call to varpart, try removing it from
the function call and moving it above as follows:

options(na.action="na.omit")
RDA_Ger <- varpart(comm, x1, x2, x3, transfo="hellinger", scale = FALSE)

Maybe that will help.

Regards,
Eric
On Sat, Sep 16, 2017 at 6:53 PM, Sophi Marmen <sofimarman at gmail.com> wrote:

            

      
      
      
    

        

      
      
    

        

  
  


  


      

    

    

      
      

        


  

        

      Bert Gunter
    

    

      
      #
    

  


  

      
Doubt that will make any difference.

My guess is that there is a function in the environment that varpart wants
to use as an object -- i.e. there is a scoping bug in varpart. But that
could be baloney, too.

Please learn about R's debugging tools. Give us the results of traceback()
after the error.

-- Bert



Bert Gunter

"The trouble with having an open mind is that people keep coming along and
sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )

        
On Sun, Sep 17, 2017 at 1:23 AM, Eric Berger <ericjberger at gmail.com> wrote: