Convention for Guinier

[DrPaulSharp]

In the Corfunc perspective and docs, the Guinier function is defined as:

$$ I(q) = e^{A+Bq^{2}} $$

whereas in the Invariant perspective and docs, we have for the Guinier:

$$ I(q) = I_{0}e^{q^{2}R_{g}^{2}/3} $$

We should choose a single convention and apply it throughout SasView.

[butlerpd]

The second is the correct equation for the Guinier which is based on a Taylor series expansion of the form factor (P(Q) around Q=0 first shown by Andre Guinier as I recall? the first equation is just a generic function and NOT the Guinier equation per se. However, there may be places where one would like to point out the simplified mathematical form of the Guinier equation? So not sure they are mutually exclusive.

That said, the way it is stated in the cofunc docs I would argue is strictly incorrect. It is defined as in the second equation above and in all scientific literature that I am familiar with. Substituting B for $R^2_g/3$ and $log(I_0)$ for A and then applying the exponential math you get the first equation which is useful when programming the math I suppose?

So I guess the bottom line is that I would agree that the corfunc doc should either replace the first equation with the second, or change the lead into to the equation to say

The scattering data is extrapolated to q=0 by fitting a Guinier function which has the general form:

[smk78]

See #1993 for an earlier discussion.

But I don't object to the proposed textual change above.