Changes in version 1.4.1                        

  - Text modifications to 'ks_test_stat' help docs, especially to note
    only tests two-sided null hypothesis.

                 Changes in version 1.4.0 (2019-07-03)                  

  - Added new function 'ks_test_stat' to more quickly calculate test
    statistic D.
  - Added new argument 'varModel' to 'LcKS' to allow users to specify
    'equal' or 'variable' variance models for mixture distributions
    (GitHub issue #6).
  - Added option to run 'LcKS' in parallel, using 'doParallel' and
    'foreach' infrastructure. Added new parameters 'parallel' and
    'cores' to control parallel behavior (GitHub issue #4).

                        Changes in version 1.3.0                        

  - Changed spline method from default to 'hyman' in qmixnorm that is
    more appropriate for handling monotonic sequences, and improves
    performance under certain mixture models. (Thanks to Qiong Zhang for
    catching the error; GitHub issue #5.)

                        Changes in version 1.2.5                        

  - Changed number of randomly generated samples per mixture components
    ('nr') to 1000, to match help file text. nr = 1000 is much faster
    and nearly equally precise (GitHub issue #1).

                 Changes in version 1.2.4 (2018-08-14)                  

  - Silenced 'verbose' arg when calling 'mclust' functions (GitHub issue
    #2).

                 Changes in version 1.2.3 (2018-03-04)                  

  - Reformatted functions to Google / Hadley style.

                        Changes in version 1.2.2                        

  - Clarify text so better explains behavior based on 'mclust' package
    functions.

                        Changes in version 1.2.1                        

  - Improved text, with corrected textual errors.

                 Changes in version 1.2.0 (2016-03-19)                  

  - Text clarifications to Help files prior to posting to CRAN.
  - Removed calculation of standard error in LcKS.

                        Changes in version 1.1.0                        

  - Steve Wang re-wrote qmixnorm to improve performance and make the
    function more compatible with behavior in other quantile functions,
    such as qnorm.
  - Improved handling of extreme probabilities in qmixnorm, based on
    algorithm by Luca Scrucca.
  - Corrected behavior when single 'sd' specified in mixture model
    functions so that results in equal-variance mixture model, and calls
    warning to alert to behavior

                        Changes in version 1.0.0                        

  - First release (to GitHub).