General
=======

All files of type .txt are ment to be read by humans and there is no strict structure in these 
files. All files of type .dat are ment to be read by computer software. These files follow
this simple structure:
* "\n" is used as the newline character (Unix-style line ending).
* One line represent one line in a matrix/vector.
* Each item in a row is separated by one space characters.

List of files
=============

Here is a list of all files included, and a short description of each file.

* README.txt ................................: You are reading me.

* statistics_and_misc_values.txt ............: Numbers from Section 12 in the paper.

* covariance_matrix.dat .....................: This file contains the covariance matrix C used in 
                                               Section 12.

* covariance_matrix_inverted.dat ............: This file contains the inverted covariance matrix 
                                               C^-1 used in Section 12.

* probability_distribution_long_double.dat ..: This file contains the probability distribution of
                                               x1 and x2 (equation 24 and 27 in Section 11). As x1 
                                               and x2 are equally distributed there is only one row
                                               in this file. The numbers are printed in scientific 
                                               form. For example 6.1034937289150548167527e-05 means
                                               6.1034937289150548167527 * 10^(-05) =
                                               0.000061034937289150548167527. There might be (very
                                               small) rounding errors due to floating point
                                               precision in C++ (long double).

* probability_distribution_times_2power32.dat: This file contains the same data as 
                                               "probability_distribution_long_double.dat". The
                                               difference being that each element is multiplied by
                                               2^32, and there is no rounding errors. You can use
                                               these numbers if you want to compute the distribution
                                               with higher precision than long double (C++) can give.

* mean.dat ..................................: This file contains the vector μ (Greek letter mu) 
                                               referenced in Section 12. This is the expected value 
                                               of R_t referenced in the paragraph before Lemma 1 in 
                                               Section 7.2, assuming x1 and x2 follows the distribution
                                               given in "probability_distribution_long_double.dat".

* variance.dat ..............................: This file contains the variance of R_t (see prev. point).

* w_vector.dat ..............................: This file contains the vector \omega defined in
                                               the 2nd to last paragraph in Section 7. This is the 
                                               result of matrix multiplication of the columns vector 
                                               in "mean.dat" and the matrix in 
                                               "covariance_matrix_inverted.dat".