5

I want to test a species' presence / absence records for spatial autocorrelation. My data contain >130,000 grids in GIS and with about 700 species' presence records.

I have read that the normal Moran's $I$ can't deal with this kind of data, but that the join count method in package spdep can do it. However, I'm new to R and I still can't understand the information and code in the help for joincount.mc or joincount.test.

My data is like this:

 gridnumber  species  
     1       1     
     2       0
     3       0
     4       1
     ……

I know how to read a shp.file into R and I know I must calculate the weight of my data, but the following steps with spdep is beyond my ability.

rafa.pereira
  • 13,251
  • 6
  • 71
  • 109
Archer
  • 63
  • 6
  • You might also want to explore the threads about `R` and `spdep` that have appeared at [gis.se]. – whuber Mar 13 '14 at 16:22
  • @gung Thank you very much. Actually getting the code is most urgent to me, for I must perform the test for a analysis at hand. However, it may be the simplest solution for my current work but a better introduction for the statistic principle would be most appreciated. Now, I think it had better to migrate my question. Thank you gung, and are you a Chinese? – Archer Mar 14 '14 at 00:56

0 Answers0