Wrong length of f_reg in evaluate_hessian

[lwitthaus]

I think the lengths of f_reg and self.reg_factor_f are not equal in

funfolding/funfolding/solution/likelihood.py

Line 261 in 2f485b0

ln_f_reg_used = np.log(f_reg * self.reg_factor_f)

if a certain number of bins is ignored in the regularization.
I believe adding the line f_reg = f_reg[self._f_slice] between

funfolding/funfolding/solution/likelihood.py

Lines 257 to 258 in 2f485b0

	if self._tau is not None:
	if self.log_f_reg:

would solve the problem, just as it is done here:

funfolding/funfolding/solution/likelihood.py

Lines 167 to 169 in 2f485b0

	if self._tau is not None:
	f_reg = f_reg[self._f_slice]
	if self.log_f_reg: