Sensitivity is defined as the response of a function to small changes in a variable. In a non-vectorial function, sensitivity can be defined as its first derivative and allows to compare the sensitivity of two points of the same function. The sensitivity of two functions can be compared by standardization, which is made in several steps. We take a reference value, calculate the sensitivity of the function near this point and standardize the result respect to this point. The sensitivity of the two functions then can be compared as they have the same range.
The four indices could be expressed as in Table C1:
|
|
|
|
|
|
Standardization of these indices around a point "x" is done as follows:
|
|
|
|
|
|
|
|
|
|
|
|
The definition of
symmetry in our manuscript can be expressed as f(u)= -f(1/u) where
u is the independent variable. When there is facilitation
and x are >1 and we use the terms
F
and xF; conversely, when there is competition
and x <1, and we use
C
and xC as 1/
and 1/x respectively. All terms range from 0+ to +
.
The above equations
can be expressed as follows (in terms of
F,
xF and
C,
xC)
|
Facilitation
( |
Competition
( |
|
|
|
|
|
|
|
|
|
|
|
|
Sensitivity is calculated
by testing the function index variation when
changes around x
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
We only allowed
for a maximum deviation of
F
and
C of
an order of magnitude from the pivotal point (i.e., small changes around x).
Replacing
= x-1,
the standardized sensitivity is expressed as
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|