Skip to content

spatial analysis with dbMEM

3 messages · Andrew Halford, Aaron Hogan, Jean Thioulouse

Original
Andrew Halford
# 
Hi Listers,

I am trying to wrap my head around spatial eigenvector analyses and have
been trying to recreate the example code attached to the dbmem{adespatial}
documentation.

I am getting an error message when I try to run the following... The data
is drawn from the oribatid dataset.

data(oribatid)

mite <- oribatid$fau
mite.xy <- oribatid$xy
mite.dbmem1 <- dbmem(mite.xy, thresh=1.012, MEM.autocor = "non-null",
silent = FALSE)
s.label(mite.xy, nb = attr(mite.dbmem1, "listw"))

Error in s.label(mite.xy, nb = attr(mite.dbmem1, "listw")) :
  unused argument (nb = attr(mite.dbmem1, "listw"))
I dont know what the nb argument is here?

# Plot maps of the first 3 dbMEM eigenfunctions
s.value(mite.xy, mite.dbmem1[,1:3])

Error in s.value(mite.xy, mite.dbmem1[, 1:3]) :
  Non equal row numbers 70 3


I checked the mite.xy and mite.dbmem1 data and they both have 70 rows?

Andy



-- 
Andrew Halford Ph.D
Research Scientist (Kimberley Marine Parks)
Dept. Parks and Wildlife
Western Australia

Ph: +61 8 9219 9795 <+61%208%209219%209795>
Mobile: +61 (0) 468 419 473 <+61%20468%20419%20473>
Aaron Hogan
# 
Hi Andy,

I am probably not the most qualified person to answer this, but I think I
may have a solution to your issue.


1.  It seems that nb is an old (or alternative) way to specify the labels
for the plot.  The documentation suggests using label.  I could not find nb
defined and was rarely used in:
https://cran.r-project.org/web/packages/adegraphics/adegraphics.pdf
<https://cran.r-project.org/web/packages/adegraphics/adegraphics.pdf>

2.  This example helped me better understand the syntax of the s.label
function:

data(mafragh, package = "ade4")
g2 <- s.label(mafragh$xy, nb = mafragh$nb, paxes.draw = FALSE)

You can see that both the $nb list and $xy dataframe have the same number
of observations (which is important- as you point out).

3.  you can extract the list from the mite data using:   attr(mite.dbmem1,
"listw")$neighbours

so putting it all together:

s.label(mite.xy, label = attr(mite.dbmem1, "listw")$neighbours)


alternatively (making the list an object named mite.nb):

mite.nb <- attr(mite.dbmem1, "listw")$neighbours
s.label(mite.xy, label = mite.nb)



Code generated the plot for me.
I hope this helps,
Best,

Aaron Hogan
(970) 485-1412
hogieskier at gmail.com

On Mon, Jan 16, 2017 at 9:50 PM, Andrew Halford <andrew.halford at gmail.com>
wrote:

  
  


  


      

    

    

      
      

        


  

        

      Jean Thioulouse
    

    

      
      #
    

  


  

      
Hi

You need to load the adegraphics package AFTER loading the ade4 and adespatial packages.

There are two versions of the s.label function : one in the ade4 package and one in the adegraphics package. The one you need here is the second one.

Jean