CHANGES: Log of significant changes between revisions

#############################################################################
# multic 0.2
#############################################################################
NEW FEATURES
  - Added in the option to use the LRT instead of the Wald statistic
    when combining the various traits (addGE).

  - Allowed the user to specify the initial values (multic.control).

  - Added a plot to show all the lods scores of the top n most influential
    families to the max LOD score (plot.family.lods).

  - Improved the error message if share.out (or possibly, mloci.out)
    does not have the expected number of lines based on the famid
    argument.

BUG FIXES
  - Calculating the residuals and collapsing on the family was crashing
    with family of size one.

  - Allowed partial matching of arguments to the residuals.multic and
    fitted.multic methods.

  - Fixed the subset argument to multic.  It was not working when the
    subset was NOT a part of the data argument.

  - get.top.n.families would crash with multiple "global peak" values.

#############################################################################
# multic 0.1.1
#############################################################################

BUG FIXES
  - Converted CLK_TCK in the C++ code to CLOCKS_PER_SEC.  CLK_TCK is
    deprecated.

  - mkdir is only defined in Splus, not in R.  multic.mkdir was
    created to encapsulate this difference.

  - Removed the "multicInput" inconsistancy in examples/calls.q.
