[R] Beginners question about Percentage similarity in R?

Richard Price freerichard at btinternet.com
Sun Sep 23 18:04:47 CEST 2007


I have been reading a paper whereby the authors took values from
Sorensons dissimilarity index and values from a percentage similarity
index and applies G-Testing to the table of values. This is carried
out to assess the differences in spider faunas (Strattton and Uetz,
1979).

I like the method but have been trying to work out what function in R
to use to get the percentage similarity. I have vegan installed and
have been using it. I think R is brilliant.  But using the formula...

PS=3D2w/(a+b)

where w=3Dminimum sum of individuals of species held in common
between the samples being compared; a and b =3 the totals of
individuals found in either region. I have had trouble working out
from the help extract from designdist which of the methods is the
percentage. 

What does it mean when someone says 'minimum sum of individuals of
species held in common between samples? 

Is it the number of species where the count is exactly the same?

So, if I collected 14 of A. curbitana from 'site a' and then 14 of A.
curbitana from 'site b' than this is a species held in common between
the samples?

In R how do I calculate this percentage similarity?

Thanks,
Richard.



More information about the R-help mailing list