Refactor the code from Manuel Rexer

Overall refactoring of the code and creation of tests that also could be used to validate other implementations.

@manuel.rexer Please give overall feedback and I'm unsure if I described the 'propagateSysUncToFreqDomain(...)' and 'calculateSysUncInFreqDomain()' correctly. So please have a closer look at the docstrings of those two functions and correct me if I have described something wrong. Also please check the README.md if you are happy with the structure and the content and if something important is missing.
Thanks.

I didn't implement the test of the data structure since It can be controlled by hand easily. But If there should be advances in the use of RDF in calculation packages/software we should come back to it and implement it.

Open TODOs:

1. The main function is quite short. Should I still leave in the short code example on how to use it? I think a real world example with a plot of a data signal would be quite cool.
2. Should I add a short documentation of the output variable 'result_struct' that the user knows what it contains and how to use it? If yes, I'm going to use a picture with some kind of directory-tree structure
3. The repository still needs a license

If you are happy with the current state you can do the merge-request by yourself.
I will open a new one for the documentation and the license.

propagateSystematicUncertainityToFrequencyDomain_package/underlying_functions/calculateSysUncInFreqDomain.m

+function [systematic_uncertainty_absolute, ...
+          systematic_uncertainty_phase] = ...
+                    calculateSysUncInFreqDomain( ...
+                                                sysunc_bias, ...
+                                                sysunc_slope, ...
+                                                sysunc_linearity, ...
+                                                sysunc_hysteresis, ...
+                                                FFT_Mean)
+% CALCULATE SYSUNCINFREQDOMAIN Calculate the systematic uncertainty in the
+% frequency domain with the mathematical expression given by the paper
+% mentioned in the README.md of this repository/package.
+%
+%   [systematic_uncertainty_absolute, ...
+%     systematic_uncertainty_phase] = ...
+%               calculateSysUncInFreqDomain( ...
+%                                           sysunc_bias, ...
+%                                           sysunc_slope, ...
+%                                           sysunc_linearity, ...
+%                                           sysunc_hysteresis, ...
+%                                           FFT_Mean)
+%
+%   -- sysunc_bias  scalar (1-by-1) real number of type double that
+%   represents the *bias* systematic uncertainty and can be obtained from
+%   the data sheet of the sensor that was used for the measurement or from 
+%   calibration protocols.
+%
+%   -- sysunc_slope  scalar (1-by-1) real number of type double that
+%   represents the *slope* systematic uncertainty and can be obtained from
+%   the data sheet of the sensor that was used for the measurement or from 
+%   calibration protocols.
+%
+%   -- sysunc_linearity  scalar (1-by-1) real number of type double that
+%   represents the *linearity* systematic uncertainty and can be *sometimes
+%   (not always given)* obtained from the data sheet of the sensor that was
+%   used for the measurement or from calibration protocols. 
+%   Use 0 if there isn't a given value for your sensor.  
+
+%   -- sysunc_hysteresis  scalar (1-by-1) real number of type double that
+%   represents the *hysteresis* systematic uncertainty and can be
+%   *sometimes (not always given)* obtained from the data sheet of the
+%   sensor that was used for the measurement or from calibration protocols.
+%   Use 0 if there isn't a given value for your sensor.
+
+%   -- FFT_Mean  FFT of the mean value vector of the "statistic
+%   oscillation".
+%
+%
+%   Returns:
+%   -- systematic_uncertainty_absolute  scalar (1-by-1) real number of type
+%   'double' that represents the absolute systematic uncertainty of the
+%   measurement signal with the given sensor in the frequency domain.
+%
+%   -- systematic_uncertainty_phase (1-by-N) vector with real numbers of
+%   type 'double' that represents the systematic uncertainty of the phase in
+%   the different frequencies of the FFT of the measurement signal with the
+%   given sensor. 
+
+
+systematic_uncertainty_absolute = sysunc_bias + sysunc_slope + (0.608 * sysunc_linearity);
+systematic_uncertainty_phase = atan(0.681 * sysunc_hysteresis ./ abs(FFT_Mean));
+
+end
+