Hola a todos!

This question may be irrelevant, but it's been annoying me for a
while. I cannot properly figure out the role of the constant "k" in
TNT's fit function, for (unless I am doing something wrong) it is not
clear to me the exact relation among the parameters.
I can see it is still a concave function (what is the most important),
but it doesn't seem to be anymore the traditional Golloboff's (1993)
"f=(k+1)/(k+1+s-m)", neither this "new" version, "f=k/(k+e)". The
values just doesn't seem to match my expectation...

For example, I expect:
k=0.01 (~clique)
e=> 0 1 2 3 4
f=> 1 0 0 0 0
;
k=1
e=> 0 1 2 3 4
f=> 1 .5 .33 .25 .8
;
k=2
e=> 0 1 2 3 4
f=> 1 .66 .5 .4 .33
*****************************************
But I get:
k=0.01 (~clique)
e=> 0 1 2 3 4
f=> 1 0 0 0 0
(this is OK);
k=1
e=> 0 1 2 3 4
f=> 1 .33 .17 .1 .07
k=2
e=> 0 1 2 3 4
f=> 1 .5 .3 .2 .14
*****************************************

And so on...
So, what is really going on? What is (or, where can I get) the real
function?

Thanks in advance!

--
Best regards,

--MSc. Peterson Lásaro Lopes
-Museu de Zoologia da Universidade de São Paulo
Avenida Nazaré, 481 - Ipiranga
Caixa Postal: 42494
CEP 04218 970
São Paulo/SP, Brasil
Tel: +55 11 6165-8123
Fax: +55 11 6165-8113
+55 11 6165-8116
http://www.usp.br/mz