Correct documentation error in Svensson zero rate formula

[twan3617]

Inconsistency was found between:

• Svensson implementation here

  QuantLib/ql/termstructures/yield/nonlinearfittingmethods.cpp

  Lines 170 to 174 in c0fac24

  	Real zeroRate = x[0] + (x[1] + x[2])*
  	(1.0 - std::exp(-kappa*t))/
  	((kappa+QL_EPSILON)*(t+QL_EPSILON)) -
  	(x[2])*std::exp(-kappa*t) +
  	x[3]* (((1.0 - std::exp(-kappa_1*t))/((kappa_1+QL_EPSILON)*(t+QL_EPSILON)))- std::exp(-kappa_1*t));

  and the
• Svensson documentation here:
  

  QuantLib/ql/termstructures/yield/nonlinearfittingmethods.hpp

  Lines 115 to 120 in c0fac24

  	\f$ d(t) = \exp^{-r t}, \f$ where the zero rate \f$r\f$ is defined as
  	\f[
  	r \equiv c_0 + (c_0 + c_1)(\frac {1 - exp^{-\kappa t}}{\kappa t})
  	- c_2exp^{ - \kappa t}
  	+ c_3{(\frac{1 - exp^{-\kappa_1 t}}{\kappa_1 t} -exp^{-\kappa_1 t})}.
  	\f]

Corrected the documentation to follow the implementation.

[boring-cyborg]

Thanks for opening this pull request! It might take a while before we look at it, so don't worry if there seems to be no feedback. We'll get to it.

[CLAassistant]

All committers have signed the CLA.

[lballabio]

Thanks for bringing attention to this—there were also a few issues in the LaTeX syntax of the formulas, I fixed those too.

[coveralls]

coverage: 72.651%. remained the same
when pulling 903dda4 on twan3617:patch-1
into c0fac24 on lballabio:master.

[boring-cyborg]

Congratulations on your first merged pull request!