2014-06-02 Gregory Belenky
* Version: 0.6.0
* General: Added new robust equivalence test.
2013-10-24 Andrew Robinson
* Version: 0.5.7
* General: updated for compatibility.
2010-01-05 Andrew Robinson
* Version: 0.5.6
* equivalence.xyplot.Rd: removed unknown macro '\bar'.
2009-01-26 Andrew Robinson
* Version: 0.5.5
* equiv.boot.Rd: removed errant space between braces.
2009-01-16 Andrew Robinson
* Version: 0.5.4
* Rd files: replaced calls to code with verb for 2.9.x.
2008-04-16 Andrew Robinson
* Version: 0.5.3
* equiv.boot.R: Explicitly checks that x and y are of equal length
and stops if not.
2008-01-10 Andrew Robinson
* Version: 0.5.2
* equiv.boot.lm.R: this function is now hidden from the user
because people were using it instead of equiv.boot(), its wrapper.
* equiv.boot.R: now provides an argument to express the interval
of indifference for the intercept in the same units as the
intercept.
2008-01-08 Andrew Robinson
* Version: 0.5.1
* equivalence.xyplot.R: corrected the computation of the
confidence interval of the intercept. Previously the interval was
computed too high. Added options to add a smooth to the plot to
aid interpretation, and to specify absolute or relative (to the
estimated mean) intervals for the intercept.
* ptte.data.R: this test no longer allows for absolute intervals
of indifference. Use tost() for such tests.
* ptte.stat.R: this test no longer allows for absolute intervals
of indifference. Use tost.stat() for such tests.
* tost.stat.R: this test no longer allows for relative intervals
of indifference. Use ptte.stat() for such tests.
2007-12-28 Andrew Robinson
* Version: 0.5.0
* tost.R: The tost.data() function has been deprecated. Please
use tost, which relies in the t.test() function and its civilized
apparatus. The advantage is that we now accommodate pooled,
unpooled, and paired data in the one function.
2007-12-16 Andrew Robinson
* Version: 0.4.4
* several: Cleaned up documentation so that internal files are hidden.
2007-11-08 Andrew Robinson
* Version: 0.4.3
* equiv.boot.R: Fixed problem in where the wrong intervals were
being used for the slope; thanks to Michael Saunders for the bug
report.