2012-03-06  V_0.0.7 Youyi Fong

    * add log.transform to rumi argument list
    * add est.conc, lb.conc, ub.conc to data frames returned by rumi

2012-03-22  V_0.0.8 Youyi Fong

    * add weighting to rumi argument list. Note that the power is -1.8 and it
    only works for non-transformed FI.
    * add drm.weights.var.power, a global variable that controls variance
    function 

2012-04-09  V_0.0.9 Youyi Fong

    * add testing vcov is null due to a numerical exception from some dataset

2012-05-11  V_0.0.10 Youyi Fong

    * add incr and auto.detect.incr at the request of Maria

2012-05-21  V_0.0.11 Youyi Fong

    * bug fix for unknown sample concentration estimation when the curve is
    decreasing
    * prompted by Maria and Devin

2012-06-24  V_0.0.12 Youyi Fong

    * change the ordering of rows in data frame, this affects the numerical
    outcome of fitted curve because drm /5pl is sensitive
    * fix a bug for plot=F & find.LOD=T
    * if a sample has multiple replicates, fi in the returned data frame now
    equals exp(Y_bar), previously it equals exp(Y_1)
    * prompted by William 

2012-08-11  V_0.0.13 Youyi Fong

    * add loq.30
    * change how the default behavior of find.lod works. Previously it finds
    95% CI by default, now it is 90%. This change is made to make it
    consistent with test.LOD, which defaults to 90%. Both defaults can be
    changed with a single parameter now, lod.ci
    * each curve will always have lod and loq. When se is bad, it will be NA

2012-08-22  V_0.0.14 Youyi Fong

    * use qt instead of qnorm for confidence interval and LODi
    testing/estimation

2012-08-30  V_0.0.15 Youyi Fong

    * change the column names of loq30 from rloq to rloq30, same for lloq
    * change the column names of lod from x.low to llodi, same for x.high
    * CI was defined on the t scale
    * LoDi was defined on the t scale, now the ci.scale.t parameter controls whether it is on t or x scale
    * LoQ was defined on the x scale using an usual definition of se
    * need to tell Devin LoDi.x=F
    * new feature in plots
    * Fix: if fi is 0 or less, things go wrong with no good error message
    (Tiffany).

2012-08-30  V_0.0.16 Youyi Fong

    * remove log.transform from the list of arguments
    * add rumi.formula, rumi.data.frame

2013-04-17  V_0.1-1 Youyi Fong, Krisztian Sebestyen 
    * Rename Ruminex to nCal
    * improve fit.jags functional interface
    * add import functions
    * add Excel plugin
    * submit to CRAN
    * paper submission
    * remove drm.weights.var.power, a global variable that controls variance

2013-05-28  V_0.2-1 Youyi Fong
    * pass stringsAsFactors =FALSE to data.frame
    * repond to referee comments, add vignette, add tests, change package
    dependencies
    * fix bug in ncal.character
    * remove incr and auto.detect.incr and handle decreasing more smartly
	* framed.sty needs to be present where kpsewhich.exe locates
	* qpdf is downloaded and path added, ghostscrip

2013-05-28  V_0.2-2 Youyi Fong
	* making changes according to major revision comments
	* add additional.plot.func to ncal list of params
	* in ncal, when plotting drm fit, having xlim sometimes messes things up,
	not sure why

2013-05-28  V_0.3-1 Youyi Fong
	* remove some checkEqualsNumeric from nCal.Rd since they fail on some
	systems (drm optimization related)
	* making changes according to minor revision comments

2013-05-28  V_0.3-2 Youyi Fong
	* restrict unit testing to x86_64, mingw32 and x86_64, linux-gnu

2013-05-28  V_0.3-3 Youyi Fong
	* after bug fix to integrate(), some tests fail. move all testing from
	function help examples to unit testing files. change tolerance for those
	checks depending on integrate() to a large number for now. See email from
	Prof. Ripley on 7/16/13

2013-05-28  V_0.3-4 Youyi Fong
	* move unit testing code for read.luminex.xls to runit.ncal.R from
	function help
	* expand unit testing to all platforms with relaxed tolerance

2013-07-24  V_0.4-1 Youyi Fong
	* in drm.fit, use default ss.fct first so that the results will match drm call in most cases
	* add check.out.of.range parameter to ncal that supports two modes of
	checks. The default mode is a new mode which resets any estimated
	concentration that is outside the range of standard concentrations


2013-11-07 Youyi Fong
	* make fi 1 in ncal if fi is 0 or negative

2013-11-29 Youyi Fong
	* use a new copy of youtil.R that does not have getVarComponent.nlme to avoid a warning after a drc update

2014-03-31 Youyi Fong
	* change } else if(all(is.na(diag(vcov.)))) { to any at Doug Lowe's
	suggestion

2014-04-23 Xuesong Yu
	* add weighted ls implementation

2014-10-31 Youyi Fong
	* remove plot.crm. need to learn callNextMethod to use it, b/c it does not
	allow plot.drc or drc:::plot.drc

2018--8-17
	* Change to use %.%
