5.0-3 (09/28/2021)
-----
Fix bug in e0.diagnose (thanks to Asmida Binti Mustafa).
Depends now on bayesTFR >= 7.0-5.
5.0-1 (04/04/2021)
-----
Support for annual simulations added.
Support for subnational projections added (e0.predict.sunat).
Updated to work with bayesTFR >= 7.0-*.
4.1-0 (10/17/2019)
-----
Defaults are now set to use the wpp2019 data.
License file added.
4.0-3 (08/23/2019)
-----
Fixed reproducibility if running run.e0.mcmc in parallel (thanks to Budi Setiawan).
4.0-2 (06/14/2019)
-----
Various arguments are now handled by options. Various internal functions changed in order to support it.
The way priors are passed to run.e0.mcmc has changed -
see ?run.e0.mcmc and ?e0options()
Support for wpp2019 added.
3.2-0 (10/13/2017)
-----
To approximate the UN settings for WPP2017, defaults changed for:
in run.e0.mcmc:
sumTriangle.lim = c(30, 86)
Triangle.prior.up=c(100, 100, 50, 100)
Triangle.c.prior.up=c(100, 100, 50, 100)
in e0.jmale.predict:
max.e0.eq1.pred = 86
Template file updated.
3.1-0 (7/27/2017)
-----
Support for wpp2017 added. wpp.year=2017 is the new default in run.e0.mcmc.
Examples re-created with new defaults.
Note two changes in country names in wpp2017:
"Czech Republic" -> "Czechia"
"Other non-specified areas" -> "China, Taiwan Province of China"
3.0-5 (05/11/2017)
-----
Restriction on sum(Delta) for initial values.
Minor fix in MCMCs.
3.0-4 (01/07/2017)
-----
Changed the way Deltas are drawn from a distribution with
a restriction on sum(Delta) (made more efficient).
Argument outliers added to run.e0.mcmc for configuring outlier ranges.
Defaults to c(-5, 10).
3.0-3 (12/02/2016) [CRAN]
-----
Added DOI to DESCRIPTION.
3.0-2 (12/01/2016)
-----
Function e0.parDL.plot added.
Argument adjusted.only added to e0.trajectories.plot.
3.0-1 (11/21/2016) [CRAN]
-----
Fixed a bug that caused run.e0.mcmc.extra not being able to pass my.locations.file (thanks to Yicheng Li).
Added functions for obtaining world and country-specific DL curves (e0.world.dlcurves, e0.country.dlcurves).
Renamed two methods due to changes in bayesTFR.
Website changed http -> https.
3.0-0 (12/4/2015) [CRAN]
-----
Added support for wpp2015 (changed imports, dependencies, defaults).
Added include_2015.
Added argument "col" and "show.legend" into e0.DLcurve.plot.
Example data re-created with wpp2015.
Fixed a bug in predict.extra for male e0.
2.2-0 (03/12/2015) [CRAN]
-----
Analog to bayesTFR, added option of overwriting UNlocations file, i.e. easily include new countries in simulations.
For this, function run.e0.mcmc and e0.jmale.predict got an argument "my.locations.file".
Fixed bug in e0.jmale.predict that caused crashing when no supplemmental data were used
(thanks to Simona Toti and Luca Mancini).
Allow estimation of the gap model for situations when no data for 1950-55 are available.
In such case, a warning is given.
2.1-3 (11/19/2014)
-----
Fixed bug in imputation of male e0 data - test added.
Fixed bug in summary of male prediction object.
Documentation to convert.e0.trajectories: added explanation about converting male trajectories.
2.1-2 (11/12/2014)
-----
In run.e0.mcmc fixed bug in passing ... into continue if iter="auto".
Fixed a bug in summary, so that it works for joint male object.
2.1-1 (6/2/2014)
-----
Time label in e0.joint.plot fixed.
2.1-0 (1/21/2014) [CRAN]
-----
Joint female-male model made consistent with Raftery, Lalic, Gerland (2014).
2.0-2 (10/10/2013)
-----
Changes in dependencies and imports.
2.0-1 (7/18/2013) [CRAN]
-----
e0.joint.plot can display observed data. Added graphical arguments obs.pch and obs.cex for the observed data.
Default wpp.year changed to 2012. wpp2012 package put into required dependencies and wpp2010 into suggested dependencies.
2.0-0 (5/23/2013) [CRAN]
-----
WPP data removed and put in separate packages (wpp2008, wpp2010). The package is now dependent on those.
1.1-6 (5/20/2013)
-----
Argument 'adjusted' added to write.projection.summary and its underlying functions.
1.1-5 (4/19/2013)
-----
e0.median.set and e0.median.shift allowed for joint male predictions.
1.1-4 (4/6/2013)
-----
e0.joint.plot: allow plotting points with identical values.
1.1-3 (4/5/2013) [CRAN]
-----
.find.package replaced by find.package.
1.1-2 (3/19/2013)
-----
Function e0.median.reset added.
Default revision number in projection summary functions set to NULL, because bayesTFR now determines it automatically.
Bug in setting starting values for Traingle and lambda fixed.
1.1-0/1 (3/13/2013)
-----
Added start.year argument into e0.predict.
Bug in e0.jmale.predict fixed (caused error when missing values in female data).
MCMCs for the Triangle parameters switched from Metropolis-Hastings to slice sampling.
1.0-2 (3/5/2013)
-----
Change due to change in bayesTFR, function for writing projection summaries - columns added.
1.0-1 (2/16/2013)
-----
In e0.dl.coverage - MAE corrected.
1.0-0 (1/31/2013)
-----
Truncation in the full conditionals of the world parameters corrected. Estimation now via Metropolis-Hastings.
Estimation of the z and lambda_z parameters via slice sampling.
Acceptance rule in MH of the lambda parameters corrected.
Default mcmc settings changed to # iterations=160000, burnin=10000, thin=225 (to have 2000 trajectories).
0.6-1 (1/3/2013)
-----
.First.lib and .Last.lib changed to .onLoad and .onUnload
0.6-0 (12/14/2012)
-----
Added support of compressed MCMC files.
Example data made smaller.
0.5-2 (11/30/2012)
-----
Minor bug fixes.
PNAS paper added to the documentation.
0.5-1 (10/20/2012)
-----
Argument old.ages.constant.gap renamed to constant.gap.eq2 and moved from e0.jmale.predict to e0.jmale.estimate.
Argument min.e0.eq2.pred added to e0.jmale.predict controlling the age when the second phase of the model kicks in.
0.5-0 (10/11/2012)
-----
Option old.ages.constant.gap added to e0.jmale.predict. If True the gap for older ages stays constant.
Bug in e0.jmale.predict fixed (related to writing out results).
Function e0.GoF.dl renamed to e0.dl.coverage. Its MSE output changed to RMSE. Added MAE and cumm.CDF.
Internal changes in *.plot.all functions.
e0.trajectories.table can return numbers for both sexes at once (added argument both.sexes).
In e0.trajectories.plot and e0.trajectories.table, added an option of outputting average female-male e0. Set both.sexes='A'.
0.4.1 (05/17/2012)
-----
Projection summary files are now automatically created also for joint male projections.
0.4-0 (05/03/2012)
-----
Function for plotting joint distribution of female and male e0 added: e0.joint.plot.
Functions added that create graphs for all countries for female-male gaps and the female-male joint distribution:
e0.gap.plot.all, e0.joint.plot.all
Prediction of joint female-male e0 is invoked automatically from the e0.predict and e0.predict.extra functions
(if the given mcmcs correspond to a female simulation).
0.3-2 (04/12/2012)
-----
Some defaults of run.e0.mcmc changed (e.g. start.year was set 1873).
Argument joint.male added to get.e0.prediction.
Added functions get.e0.jmale.prediction and has.e0.jmale.prediction.
Function e0.trajectories.plot can plot distribution for both sexes into one plot (argument both.sexes).
Example data reduced from 3x50/2 iterations to 2x60/2 iterations.
Functionality for overwriting country specific prior parameters added (argument country.overwrites in run.e0.mcmc and run.e0.mcmc.extra).
Arguments verbose.iter and constant.variance added to run.e0.mcmc.
e0.coda.mcmc.list renamed to e0.coda.list.mcmc to be consistent with bayesTFR's function coda.list.mcmc.
Some clean up of the experimental code.
0.3-1 (02/08/2012)
-----
The GoF.dl function renamed to e0.GoF.dl and its implementation was consolidated with tfr.GoF.dl in bayesTFR.
Arguments added to run.e0.mcmc for setting the lower and upper bounds of truncated normal prior distributions.
A few arguments added to e0.trajectories.plot.
New functionality for handling supplemental historical data implemented.
New functionality for joint prediction of female and male (Lalic et al) implemented.
Some experimental functions added.
0.3-0 (11/17/2011)
-----
Performance improved in two ways:
1. MCMCs are converted to environments (instead of being lists), so that they are not copied every time.
2. A few functions implemented in C.
Function GoF.dl for computing goodness of fit added.
An argument 'predictive.distr' added to the e0.DLcurve.plot to plot predictive distribution of the DL function.
A bug in the e0.DLcurve.plot function fixed that made the quantile curves noisy.
0.2-1 (10/13/2011)
-----
A few internal modifications.
Function get.countries.table added.
0.2-0 (05/06/2011)
-----
Official UN WPP 2010 data added to the package and the default for argument wpp.year set to 2010.
Argument 'gender' renamed to 'sex'.
Feature of auto-simulation implemented in run.e0.mcmc.
Argument gender in run.e0.mcmc changed to c('Male', 'Female').
Feature of 'auto-simulation' added to run.e0.mcmc and continue.e0.mcmc.
0.1-1 (03/11/2011)
-----
Function e0.map.gvis added.
Method get.mcmc.meta put into NAMESPACE.
0.1-0 (01/19/2011)
-----
Initial CRAN version.