Version 1.2
-------------

- the SWEEP-operator has been introduced as C-implementation in two versions:
  1) An adapted, speed-optimized version for computing ANOVA Type-1 sum of squares and corresponding degrees of freedom.
     This has been set as default method. This speeds up fitting models with many effects (random or fixed).
  2) A full SWEEP-operator for matrix-inversion ('Sinv') replacing calls to function 'MPinv' speeding up the process
     of obtaining a generalized inverse.
- one can now choose between the "qf" and the "sweep" method in functions 'anovaVCA' and 'anovaMM', where the former
  corresponds to the approach based on finding A-matrices for expression ANOVA Type-1 sum of squares as quadratic form in y,
  which is a prerequisite of applying the exact method for finding the variance-covariance matrix of variance components (VarVC.method="scm")
- in function 'anovaVCA' time consuming step of solving mixed model equations is now optional (argument "solveMME")
  since these are not required for the core of a VCA-analysis. Whenever this step is a prerequisite for other computations
  function 'solveMME' will be called and the original object will be updated automatically.
- function 'solveMME' was also refactored, outsourcing computation of the Q-matrix to function 'residuals.VCA',
  which is only needed in computing studentized, conditional residuals
- in function 'varPlot' now vertical lines can be drawn via argument 'VLine' as additional option to highlight
  levels of one or multiple factor variables besides argument 'BG'
- in function 'varPlot' argument "max.level" was introduced setting an upper limit for the number of factor-levels for which 
  in the table indicating the experimental design vertical lines will be added, if this limit is exceeded, the number of levels
  will be printed instead as e.g. "N=26"
- in function 'varPlot' graphical "xaxs" and "yaxs" are not reset to their values prior to calling this function, avoiding 
  annoying clipping of e.g. horizontal lines, which never reached the upper end of the plotting region
- function 'Solve' was added, using the Cholesky-root for solving a system of linear equations, respectively, inverting
  a matrix
- dataset "dataEP05A2_example" has been renamed to "Glucose" the original name in the CLSI EP05-A3 guideline
- dataset "CA19_9" has been added, a multi-sample, reproducibility dataset from the CLSI EP05-A3 guideline
- now using Diagonal(N) instead of Matrix(diag(N)) which was a remainder from early phases of development and terribly slow
- by-processing is now supported in functions 'anovaVCA', 'anovaMM', 'VCAinference' and 'print.VCAinference'
