THIS IS THE CHANGELOG OF THE "embryogrowth" PACKAGE
Please note that only the most significant changes are reported here.
All the major changes are indicated inside the R code.
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.