[R-sig-eco] keeping a variable as qualitative even if it is coded with numbers

[Karen Kotschy]

Hi Amelie

I am not familiar with the workings of fourthcorner analysis, but suspect 
that it involves calculation of dissimilarities. In this case, the way 
binary variables like yours are treated can be very important.

Have a look at the function dist.ktab in the package ade4 and the 
reference on the help page:

Pavoine S., Vallet, J., Dufour, A.-B., Gachet, S. and Daniel, H. (2009) On 
the challenge of treating various types of variables: Application for 
improving the measurement of functional diversity. Oikos, 118, 391-402. 

You should probably code your variables as "fuzzy" (if a species can 
belong to more than one group at a time e.g. wind and water, short and 
long), or otherwise as "binary multichoice".

I'm not sure, however, whether your fourth corner analysis can use the 
output from dist.ktab as is. But it would be important to find out exactly 
how binary variables are dealt in the fourth corner analysis, and to code 
them appropriately. For example, are they handled as symmetric or 
asymmetric binary variables? This makes a difference to the calculation of 
dissimilarty values.

Regards
Karen

---
Karen Kotschy
Centre for Water in the Environment
University of the Witwatersrand, Johannesburg, South Africa


 
On Fri 26Nov10, amelie_can wrote: > 
> Hello, 
> 
> I am running a fourthcorner analysis which links species abundance,
> functionnal traits and environmental variable. One of my trait is about seed
> dispersal (qualitative variable) in which I have four classes: wind, water,
> short and long distance. One species can have many types of dispersal. So in
> my original table, I created four different column. 
> 
> ex:  
> 
>        wind         water        short         long
> sp1    1               1              0             0
> sp2    0               0              1             0
> sp3    0               1              1             1
> 
> and so forth. My problem is that those variable are considered as
> quantitative but they are qualitative. 
> I tried to transform them using the function as.factor but then it includes
> the 0 in my analysis and I am not interest in that. 
> 
> Any ideas to make those variable "qualitative"?
> 
> Thank you all, 
> 
> Amelie
> -- 
> View this message in context: http://r-sig-ecology.471788.n2.nabble.com/keeping-a-variable-as-qualitative-even-if-it-is-coded-with-numbers-tp5778025p5778025.html
> Sent from the r-sig-ecology mailing list archive at Nabble.com.
> 
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology