THIS IS THE CHANGELOG OF THE "embryogrowth" PACKAGE

          CHANGES IN VERSION 6.3    (2016-07-17)

* Minor update of plot.NestResult() to show TSP at the top of temperatures - 6.2.1
* Caretta caretta stages are no more defaults for plot.NestResult() - 6.2.1
* equation Hulin for tsd() is more robust - 6.2.2
* fixedparameter is used for mcmc - 6.2.2
* A matrix for SE can be used in plotR() for anchored parameters to use quantiles - 6.2.3
* New parameter x.SE to control the envelope of R - 6.2.3
* cex.axis and cex.lab works for temperature scale in plotR_hist - 6.2.4
* tsd indicates durations or temperatures correctly depending on data input - 6.2.5
* plot.tsd indicates durations or temperatures correctly - 6.2.6
* It is possible to show anchors in plotR() and plotR_hist() - 6.2.7
* New columns and information in STSR_TSD database - 6.2.8
* STRN() has been changed to be more reliable if the order of nests is changed - 6.2.8
* Bug correction in info.nests() - 6.2.9
* New option to not show observations in plot.tsd() - 6.2.9
* New parameter parallel for info.nests() and GTRN_MHmcmc() to control the use of parallel version within R (preparation for SparkR version) - 6.2.10
* batchsize option has been removed from GTRN_MHmcmc() - 6.2.10
* New example for movingincubation() - 6.2.11
* Introduction of Weibull function to fit thermal reaction norm - 6.2.11
* New parameter metabolic.heating in info.nests() - 6.2.11
* New parameter metabolic.heating in movingincubation() - 6.2.12
* range.CI parameter of predict.tsd() is replaced by quantiles - 6.2.13
* New temperature.heterogeneity parameter in info.nests() and movingincubation() - 6.2.14
* Changes in documentation - 6.2.15
* Better managment of error in SE during SearchR() - 6.2.16
* New function SEforR() to estimate SE for a set of parameters based on Fisher information matrix - 6.2.16
* Estimation of anchored model is more rapid - 6.2.16
* New parameter XlimSE for plotR and plotR is simplified - 6.2.17
* New model for anchors (polynomial fit) - 6.2.18
* New parameter hessian to return the hessian matrix in SEforR() - 6.2.18
* New model for anchors (linear fit) - 6.2.19
* Correct a bug in STRN() that prevent SE estimation - 6.2.19
* Object returns by STRN() is more complete and can be used by compare_AIC - 6.2.20
* info.nests() can use an object created by STRN() for the parameter SexualisationTRN - 6.2.20
* Call to MCMC is simplified - 6.2.21
* info.nests() use a parameter x rather than parameters - 6.2.21
* info.nests() can use both x and parameters to provide back compatibility - 6.2.22
* plotR() and plotR_hist() return a list with confidence intevals - 6.2.23
* Correct a bug in info.nests() (Thanks to Taylor R. Roberge to identifiy it) - 6.2.24
* xlimR of plotR() or plotR_hist() can be a vector of values - 6.2.24
* New example for plotR() using xlimR as a vector - 6.2.25
* shiny version has been removed as it was not maintained anymore - 6.3

          CHANGES IN VERSION 6.2    (2016-01-08)

* Update of documentation

          CHANGES IN VERSION 6.1.3-6.1.5    (2015-12-04)

* plot.NestResult() and plotR_hist() can be used with layout() or mfrow() - 6.1.3
* New database stages for embryonics stages, mass and SCL. Will be expanded in next versions - 6.1.4
* Solve a problem with mcmc - 6.1.5

          CHANGES IN VERSION 6.1.2    (2015-10-22)

* Correct a problem with info.nest when timeseries are of different length during resampling of parameters
* Correct a problem with info.nest when fill is not NULL

          CHANGES IN VERSION 6.1.1    (2015-10-16)

* Correct a problem occuring only with windows version

          CHANGES IN VERSION 6.1    (2015-10-11)

* Correction of movingincubation() to work with the new version of info.nests()
* In info.nests(), when temperatures are changed, if all NestResult$test parameters are the same and test is NULL, the test parameter is taken from NestResult$test
* metric.end.incubation in info.nests() and plot.NestsResult() is recycled if necessary
* plot.NestsResult() uses info.nests(), new parameter show.metric
* if metric.end.incubation is NA, the embryo metrics for TSP are scaled by metric of hatchlings at the end of incubation
* if metric.end.incubation is NULL, the embryo metrics for TSP are scaled by the maximum metric of mean in test parameters
* if metric.end.incubation is a value, it is used directly for scale
* New data for info.nests(): MiddleThird.MassWeighted.temperature.mean and .se
* Correct a bug for info.nests() if there is only one temperature
* Correct a bug for searchR() that prevent return a result in rare occasions
* metric and summary can be used when partial incubations are observed using metric.end.incubation=x with x being the final size
* Correct a bug for fill parameter in info.nests()
* New data for embryological stages: Emys orbicularis.mass and Emys obicularis.SCL. They are store in TSP.list
* Rename Caretta caretta embryological stages: Caretta caretta.SCL
* After plotR_hist(), the y-scale is the scale for R graph (previously it was the one for histogram)
* Only the function lsoda is imported from deSolve to prevent a warning during check

          CHANGES IN VERSION 6.0.1    (2015-06-30)

* Change info.nests() for R 3.2 compatibility
* Minor changes in FormatNests(); test if names are unique
* Correct a bug for Hill model of TSD
* Add optimx as an option for searchR()
* The latest version can always been installed using:
  install.packages("http://www.ese.u-psud.fr/epc/conservation/CRAN/embryogrowth.tar.gz",
    repos=NULL, type="source")
* $SD is store also in mcmc objects. This SD is the SE for points estimates.

          CHANGES IN VERSION 6.0    (2015-04-15)

* FormatNests() can convert old format database using data=xxx with xxx being the old format database
* info.nests() is much more complete
* polynom package is now in dependencies
* New columns IndiceK in FormatNests() output to accelerate SearchR()
* predict.tsd() uses the new model for CI estimation
* plot.tsd() does not make any estimation; it just plot the object created by tsd()
* New method for confidence interval in tsd()
* If stopattest=TRUE is used in .fonctionfit() the time for end of incubation is estimated using polynomial interpolation
* Hmisc package is no more required
* new parameters SE and fixed.parameters in tsd()
* The anchored parameters can use Celcius degrees
* Many changes in .fonctionfit .STRNfit (hidden fonctions)
* searchR() is more rational for initial tests of weights and better manage parallel runs
* Add a progression bar for predict.tsd()
* MHmcmcAlgo(), minmax.periodic(), temperature.periodic() and info.sun() functions are transfered to new package: HelpersMG
* STRN() does not use replicate.CI for info.nests()
* Add parameters= for GenerateAnchor() and GRTRN_MHmcmc_p()
* The weight scheme for CTE called weigthed mean temperature during TSP can use another thermal reaction norm: the thermal reaction norm for sexualisation (STRN). The new CTE is named TSD.mean.temperature
* New CTE named TSD.weighted.mean.temperature. It includes both the thernal reaction norm for growth rate and the thermal reaction norm for sexualisation.
* New function STRN() to fit the Sexualisation thermal reaction norm
* New function STRN_MHmcmc() to model the enveloppe of the STRN
* New parameters to continue MCMC run from previous run in GRTRN_MHmcmc(), STRN_MHmcmc() and tsd_MHmcmc()
* Correct a bug to estimate the mean temperature within the TSP
* If weight is included in FormatNests, it is used in searchR()
* Correct a bug when plotR() is called without result parameter
* Correct a bug for info.nest() with SE
* Temperatures or durations can be associated with SE for predict.tsd()
* TestParallel() function display the user time of system.time()
* Package polynom is required for anchored model
* Anchored model uses now Lagrange polynoms rather than loess fitting method
* When anchored model is negative, it is fixed to 0
* New function sun.info()
* minmax.periodic() can use different minimum or maximum time each day. Correct a bug for SD in minmax.periodic

          CHANGES IN VERSION 5.1    (2014-10-05)

* New version of STRES_TSD and help for this database is more clear
* Future changes in STRES_TSD will not be longer be reported here
* tsd() can now uses also incubation period
* Equation Hulin (from Hulin et al 2009) is implemented in tsd()
* Equation Double-Richards (from Girondot, submited) is implemented in tsd()
* Packages numDeriv and entropy are now only suggested
* New parameters lty.temperatures and lwd.temperatures for plot.NestsResult()
* Predict.tsd() without temperatures or durations uses original temperatures or durations
* New function plot.tsd() 
* New possibility to use Bayesian estimate for tsd()
* Remove parallel computing for windows in MCMC
* New functions minmax.periodic() and temperature.periodic()

          CHANGES IN VERSION 5.0    (2014-08-15)

* New bibliographic references and first version in CRAN
* Remove parallel computing option for Windows
* Remove local update

          CHANGES IN VERSION 4.14    (2014-06-14)

* Correct bug for Richards equation in tsd and predict.tsd() works faster

          CHANGES IN VERSION 4.09-4.13    (2014-06-11)

* New function logLik() applied to NestsResult and tsd
* New version of STRES_TSD
* No return line from plot()
* x-axis is returned correctly after plotR_hist()

          CHANGES IN VERSION 4.05-4.08    (2014-06-03)

* predict.tsd() is more rapid and can be used without CI
* New version of STRES_TSD
* remove dependency of many packages
* new parameter lwd and lwdCI for plotR() and plotR_hist() 

          CHANGES IN VERSION 4.04    (2014-04-14)

* Correct a warning in plotR() when Scale is used
* Correct a bug in plot.nestresult() when for special length of time series

          CHANGES IN VERSION 4.01-4.03    (2014-03-05)

* 1000 iterations are allowed for tsd() and test if convergence is reached
* new way to define initial P value for tsd()
* new automatic xlim for tsd() function to show all TRT 
* many changes in shiny code. Remove visibility of ShinyEmbryogrowth()

          CHANGES IN VERSION 4.00    (2014-03-02)

* shiny application is included. Call it using ShinyEmbryogrowth()

          CHANGES IN VERSION 3.02    (2014-03-01)

* mcmc results for anchored parameters use 10000 iterations
* Correct a warning when there is no variability in TSP

          CHANGES IN VERSION 3.01    (2014-02-12)

* Possible to save intermediate results during fitting

          CHANGES IN VERSION 3.00    (2014-02-08)

* Possibility to define R by anchor of loess fitting

          CHANGES IN VERSION 2.30    (2014-02-02)

* New version with corrected rK value
* weight parameter is store in result of searchR() and used for mcmc
* Add the function switch.transition()
* New function merge.mcmcComposite() to merge two objects mcmcComposite

          CHANGES IN VERSION 2.28    (2014-01-24)

* Correct a bug in movingincubation()

          CHANGES IN VERSION 2.27    (2014-01-23)

* Correct a bug in movingincubation()
* Predict method for tsd

          CHANGES IN VERSION 2.26    (2014-01-22)

* Add SE for weight temperature by growth during TSP and mean temperature during TSP
* Add SE for borders of TSP
* Store weighted temperature by growth during TSP for movingincubation()

          CHANGES IN VERSION 2.25    (2014-01-21)

* Calculate mean temperature in middle-third of incubation and weight temperature by growth during TSP

          CHANGES IN VERSION 2.24    (2014-01-16)

* New function to change from 4p to 6p parameters functions and reverse

          CHANGES IN VERSION 2.23    (2014-01-08)

* Correct a bug for linear or exponential derivative and parallel computing
* Add mar parameter for plot.NestResults()

           CHANGES IN VERSION 2.22    (2013-12-22)

* TRT is calculated for Hill and Richards equations in tsd() function
* Limits and their standard errors are calculated in tsd()

           CHANGES IN VERSION 2.21    (2013-11-22)

* GSD, Hill and Richards equations are included in tsd() function
* Many improvments for tsd() function

           CHANGES IN VERSION 2.19-2.20    (2013-09-26)

* In MovingIncubation(), a timesseries at the end of incubation can be provided to model the metabolic heating
* SE of parameters are used if replicate.CI is different from 0

           CHANGES IN VERSION 2.18    (2013-09-09)

* Correction of bug in generateIL.df and info.nests
* generateIL.df renamed GenerateConstInc
* New parameter xlimR to show only part of R function in plotR and plotR_hist

           CHANGES IN VERSION 2.17    (2013-08-20)

* hist.nests and hist.nestsresults return a list with the complete set of temperatures

           CHANGES IN VERSION 2.16    (2013-07-01)

* New function generateIL.df

           CHANGES IN VERSION 2.15    (2013-06-25)

* New parameter OutPlot for plot.NestsResult()

           CHANGES IN VERSION 2.15    (2013-06-24)

* Correct a bug for mean temperature of a nest

           CHANGES IN VERSION 2.14    (2013-06-23)

* New version of uploading to ESE server after building package
* New parameter col for Entropy
* Output of tsd() takes into account that the regressor could be not the temperature

           CHANGES IN VERSION 2.11    (2013-06-11)

* Better format for examples

           CHANGES IN VERSION 2.10    (2013-04-08)
           
* New parameter for embryogrowth_MHmcmc_p() to automatically accept proposed priors
* New parameter echo for likelihoodR()

           CHANGES IN VERSION 2.09    (2013-04-04)
           
* Correct a bug in mean.temperature and mean.temperature.TSP in info.nest() and plot()
* Add a parameter to plotR() to manage where to display information "mean" and "confidence interval"

           CHANGES IN VERSION 2.08    (2013-04-04)
           
* Correct a bug in plotR_hist() if parameters for plotR() were used
* Add the progress parameter in info.nests
* Calculte the mean temperature during all incubation in info.nest() and plot()

           CHANGES IN VERSION 2.07    (2013-02-21)
           
* Parameter weight is added in FormatNests(), searchR() and likelihoodR()
* Correct the test in SearchR() to be sure that fixed and fitted parameters are not duplicated
* New function weightmaxentropy() which searches for the relative weights for each nest to maximize entropy
* New database fitted with weight

            CHANGES IN VERSION 2.06    (2013-02-15)
           
* Correct Movingincubation() to take into account the _L list return by .SSM()
* Correct plot.NestsResult() for truncated numbers
* as.par.mcmcComposite() and as.mcmc.mcmcComposite() are transfered to the package phenology.
* Loading of package does not failed if the description is not available online.
* Movingincubation() returns a dataframe with time and length of TSP

            CHANGES IN VERSION 2.05    (2013-01-27)
           
* many minor changes in plot.NestsResult for better ylimS managment and to prevent an error when no SD is available

           CHANGES IN VERSION 2.04    (2013-01-09)
           
* tsd() show the complete curve depending on xlim parameter
* plot.NestsResult returns a list with two objects: summary with the summary statistics for each nest and traces with the change of length or mass for each nest

           CHANGES IN VERSION 2.00-2.03    (2013-01-02)
           
* Embryo growth can be modeled using two sets of equations, one for early development and one for late development
* as.mcmc() can be used with mcmcComposite objects
* library zoo is required to use na.locf()
* Add the function tsd() to fit data of temperature-dependent sex determination
* Correction of bug for show.box and ylimH parameters in plotR() and plotR_hist()

           CHANGES IN VERSION 1.23    (2012-11-11)

* New parameter for plot(NestsResult): ref.stage=TRUE ou FALSE
* Modeled size at the end of incubation and likelihood for each nest are returned by plot()

           CHANGES IN VERSION 1.22    (2012-11-09)

* New function MovingIncubation() to run the model day by day on a timeseries
* summary() for Nests objects
* Add simplify=TRUE for FormatNests()
* mcmc data for 6p: result_mcmc_6p_80
* Correct scale of x axis for histogram and add legend for Y axis in plotR_hist()
* Legends are not drawn twice in plotR() with lists
* Initial point is taken into account for as.par.mcmcComposite()

           CHANGES IN VERSION 1.20-1.21    (2012-11-01)

* Update can use install_url() from package devtools
* New function GenerateTest(series, size, previous) to generate test parameter for searchR()
* Color of embryo mass in plot can be changed using a vector of colors

           CHANGES IN VERSION 1.14-1.19    (2012-10-31)

* Remove the use of .EGR.env environment for searchR() in 1.14 and in plot() in 1.17
* Package can load when no internet connection is available
* Add a function testParallel() to evaluate if the option parallel is interesting for the running system
* Add class mcmcComposite to result of embryogrowth_MHmcmc()
* Add BatchSE and TimeSeriesSE elements of the output of embryogrowth_MHmcmc()
* New function plotR_hist() to display both RGraph and hist of temperatures
* Rename as_mcmc() to as.mcmc2.mcmcComposite()
* New function as.par.mcmcComposite() to get parameters at maximum likelihood from mcmcComposite object
* Add a parameter force=TRUE or FALSE to force update

           CHANGES IN VERSION 1.11-1.13     (2012-10-23)

* Rename extractMCMC() to as_mcmc()
* Correct a bias when the temperature reaches a maximum or minimum
* Allow parallel computing to be disabled

           CHANGES IN VERSION 1.10      (2012-10-21)

* Uses parallel computing for searchR() and embryogrowth_MHmcmc()

           CHANGES IN VERSION 1.09      (2012-10-19)

* New functions: extractMCMC() to be used with mcmcComposite objects to get an object to be use with coda package
* New functions: embryogrowthUpdate() to update package
* Package version is checked at package loading
* Add the defaults plot parameters for plotR()
* Possible to plot several growth rates using list objects
* Correct a bias when the temperatures always increase or decrease during incubation

           CHANGES IN VERSION 1.08      (2012-10-07)

* Use mcmc methods from phenology package

           CHANGES IN VERSION 1.07      (2012-10-06)

* Remove plotnest() and plotMHmcmc() function to keep only plot().
* If optimization is not finished after maxiter, it shows the value of fitted parameters and continues.
* Change mcmcNests class to mcmcComposite.
* plot.Nests() returns also the mean temperature within the TSP.
* The package is renamed embryogrowth
* The objects wihtin the mcmcComposite object are mcmc and mcmc.list objects
* Add a method summary of mcmcComposite object
 
 CHANGES IN VERSION 1.04      (2012-05-28 -> 2012-07-28)

* Add the possibility to use fixed parameters
* Add the three new class: mcmcNests, Nests and NestsResult
* Add Bayesian MCMC script and plot of MCMC
* Many minor changes for speed improvment
* Some bug corrections
* Change T12H for 6-parameters equation to DT
* Catch the error of singularity of matrix
* Better managment of test size if the temperatures are changed

           CHANGES IN VERSION 1.05      (2012-07-30)

* The MHMCMC algorithm can now be used by other packages.
* hist.NestsResult uses the same script as hist.Nests
* plot gives information on the position of first and second third and Sum of Square difference between these and the position of stages 21 and 26.

            CHANGES IN VERSION 1.00-1.03 (2012-05-27)

* first versions as a package. New functions added sequentially.








