10.0-0/1 (08/09/2023)
------
Making default datasets from wpp2022 work with pop.predict.
pop.pyramid and pop.trajectories.pyramid got an argument for showing birth years, called show.birth.year.
pop.trajectories.plot: default for sum.over.ages changed to TRUE
Added argument xshift to pop.trajectories.plot and pop.trajectories.table.
New mapping function, pop.ggmap().
Argument allow.negative.adj can be passed to write.pop.projection.summary() controlling
if adjustments can yield negative results.
Option of giving migration inputs as rates.
Fixed bug related to extracting observed migration which did not work for aggregations
(thanks to Crystal Yu).
Options for additional settings in the mortality pattern file (vwBaseYear):
SmoothDFLatestAgeMortalityPattern (degree of freedom for smoothing)
LatestAgeMortalityPattern can have 2 values (a negative for exclusion of latest time points and a positive for inclusion)
AgeMortProjMethod1/2 can have value modPMD which uses a modified version of PMD (modPMD in MortCast)
Added support for UN-based disaggregation into migration schedules (argument mig.age.method = "un" in pop.predict)
9.0-0/1 (11/28/2021)
-------
Support for 1x1 projections (1 year age x 1 year time); still an experimental feature!
- argument "annual" in pop.predict
- age in expressions goes from 0 to 100 (observed) or to 130 (projected)
- child-bearing ages expanded to 10-54 (only for 1x1 projections)
- for subnational projections not yet supported!
The cohort component method modified so that it exactly reproduces the UN-implemented CCM (thanks to Sara Hertog for help).
A 5x5 projection can be generated using annual TFR and e0 inputs. They are hen averaged.
Fixed bug in write.pop.projection.summary for "tpopage" (reported by Budi Setiawan).
The inputs argument in pop.predict and pop.predict.subnat accepts total migration (mig) or sex-specific total migration (migMt, migFt). The disaggregation into ages is done via Rogers-Castro.
Another new element of inputs argument in pop.predict and pop.predict.subnat is GQpopM and GQpopF which can be used to define group quarters, i.e. a part of the population that should be excluded from the cohort component method.
Support for the "psr" option in various functions removed as it can be derived using expressions.
Other mortality projection methods than Lee-Carter can be used in subnational projections (argument lc.for.all enabled).
Updates in bayesTFR now allow to define the argument "country" in various functions as ISO-2 or ISO-3 characters (pop.trajectories.plot, pop.pyramid, summary, etc)
New functions:
o project.pasfr - allows to generate PASFR outside of a bayesPop simulation
o migration.totals2age - disaggregates migration into ages
o mac.expression1 - generates an expression for the mean age of childbearing for a 1x1 projection
o rcastro.schedule - returns the default Rogers-Castro age distribution
8.1-3 (5/5/2020)
-----
Minor updates related to changes in MortCast.
8.1-2 (3/10/2020)
-----
Make get.survival function work also for one time period (reported by Budi Setiawan)
8.1-1 (10/25/2019)
-----
Defaults set to use data from wpp2019.
Restoring graphical pars in various plot functions
(thanks to Ludger Sandig for the patch)
Added new expression functions:
mid.period - for population in the middle of time periods
period.diff - difference between the end and beginning of time periods
period.ratio - ratio of the end and beginning of time periods (e.g. annual growth rate: 100 * log(period.ratio(PQA))/5)
Removed dependency on the DemoTools package.
8.0-0 (06/18/2019)
-----
Support for wpp2019.
New expression "E" for the ex column of the life table (life expectancy at age x).
pop.predict can run in parallel.
Age-specific mortality is now computed using MortCast functions.
Added mortality methods for PMD, HIVmortmod, LogQuad and blends of any two methods. Arguments lc.for.all and lc.for.hiv control if Lee-Carter is used (TRUE) or methods set in the vwBaseYear dataset. Added dependency on HIV.LifeTables.
Added aggregation of mortality rates. Added dependency on the DemoTools package.
Various bug fixes.
7.0-1 (04/02/2019)
-----
Fixed bug in writing summary for cases when projection is generated
for one country only (thanks to Anna Janicka)
7.0-0 (10/11/2018)
-----
Support for subnational population projections:
- functions pop.predict.subnat & pop.aggregate.subnat
- added subnational data for Canada
6.2-4 (03/15/2018)
-----
Fixed a bug that caused the projection to fail when the option
stringsAsFactors was set to FALSE (reported by Jim Oeppen).
Fixed an issue with PASFR when projecting beyond 2100.
6.2-3 (10/22/2017)
-----
Replace NA values of historical population with 0.
(In wpp2017 the open age group for years before 1990 is 85+ and thus the pop datasets are NA for old ages. This was problematic in various sum and plotting functions.)
6.2-2 (10/11/2017)
-----
Fixed bug in adjusting function.
Unit test for adjustments.
6.2-0/1 (09/29/2017)
------
Changes needed to work with wpp2017, including setting new default for wpp.year.
6.1-1 (08/09/2017)
-----
Accent syntax changed in CITATION (request from CRAN).
6.1-0 (04/21/2017)
-----
Added function project.pasfr for projecting PASFR outside of pop.predict (thanks to Igor Ribeiro).
6.0-5
-----
Truncation of life tables to a given open age group (thanks to Patrick Gerland) --> Function
LifeTableMx has new arguments open.age and radix.
6.0-4 (02/10/2017)
-----
Improvements in deriving age-specific migration.
6.0-3 (11/29/2016)
-----
Added CITATION file and updated references with bayesPop JSS paper.
Function get.trajectories.close.to renamed to extract.trajectories.eq; added functions
extract.trajectories.ge and extract.trajectories.le.
6.0-2 (11/21/2016)
-----
Added arguments to age.specific.migration (rescale, ages.to.zero).
Fixed bug in age.specific.migration that caused using wrong country names.
Fixed bug in pop.combine.
Fixed a small bug in adjustments.
6.0-1 (01/29/2016)
-----
Option of collapsed life table added to the function LifeTableMx (argument include01).
Bug found by Marta Anacka fixed (matrix collapsed to vector if only one year of observed data given).
6.0-0 (01/15/2016)
-----
Added support for wpp2015.
Life table functions revisited. (Thomas Buettner)
Cohort deaths converted to period deaths. (Thomas Buettner)
Bug in present mortality rates fixed.
Added functions get.pop.ex and get.pop.exba for retrieving results of expressions.
Added a function for linking given population trajectories to the probabilistic
input components (migration, TFR, e0). Called "get.trajectories.close.to".
5.4-0 (7/17/2014)
-----
A few fixes in adjustments.
Added function for creating an expression for the mean age of child-bearing, "mac.expression(country)".
pop.pyramidAll got a logical argument "one.file" for plotting everyhting into one file.
pop.trajectories.pyramid accepts now the "indicator" argument (analogous to pop.pyramid)
to plot deaths and births.
Plotted trajectories are shown in a transparent color. Default colors of pop.pyramid
changed (thanks to Tim Riffe).
Argument my.locations.file added to pop.predict to allow projections for other
than the UN countries/regions (thanks to Filippo Temporin).
Aggregations for ASFR and PASFR implemented.
Added PASFR to write.pop.projection.summary.
Added functions for extracting and plotting cohort data
(functions cohorts, pop.cohorts.plot)
Added functions pop.trajectories and get.trajectory.indices.
5.3-4 (4/10/2015)
-----
Projecting age-specific mortality:
- dealt with jump-off bias - different way of computing a_x (country specific)
- implemented rotation of b_x (Li, Gerland 2013)
- some countries use modeled b_x
Implemented projection of percent age-specific fertility rate (Kantorova 2014)
Added 'R' into expressions for obtaining percent age-specific fertility.
plot.by.age got a few additional graphical arguments.
5.2-2 (2/4/2015)
-----
Fix in litem function - made compatible with R-devel.
5.2-1 (11/12/2014)
-----
Assuring population age groups are not negative (adjusting migration).
Bug in computing deaths for observed data fixed.
5.2-0 (10/3/2014)
-----
Added an option of using projections of mx only, without calculating e0.
Fix in pop.pyramid - the argument pop.max now works.
5.1-5 (8/13/2014)
-----
Minor fixes.
5.1-1 (5/9/2014)
-----
Time points for population indicators changed to end of time periods instead of the middle year.
Allowing adjustments to wpp2012 (in output functions as well as in pop.aggregate)
5.0-0 (4/28/2014)
-----
Allows probabilistic migration: trajectories of net counts can be passed in an csv file to pop.predict
(no re-balancing implemented yet).
If keep.vital.events is TRUE, the migration trajectories are stored with all the other country-specific data
and can be accessed using the "G" expression.
Aggregation 'method's changed to 'input.type' in pop.aggregate. Default input types are now
'country' (former independence method) and 'region' (former regional method).
Thus, default names of aggregation directories are aggregations_country and aggregations_region.
Updated dataset on migration type and year (vwBaseYear2012) [Patrick Gerland].
Stored bayesPop.prediction object slightly changed.
A column 'Trajectory' is now required for trajectories files of TFR, e0 and migration (if used).
In pop.predict the input elements [tfr|e0F|e0M].file accept the keyword "median_" to use the median WPP projection.
(Note that if nothing is given it takes the median, low and high as default.)
4.3-0 (4/4/2014)
-----
pop.aggregate can handle any number of overlapping aggregations in the location file.
4.2-2 (12/2/2013)
-----
Function get.pop exposed to outside.
4.2-0/1 (10/10/2013)
-----
New expression function - pop.combine.
Changes in dependencies and imports.
4.1-0/1 (7/22/2013)
-----
Setting 2012 as new default for wpp year. Adding wpp2012 into dependencies.
Default input e0 and TFR in pop.predict are taken from the wpp package as three trajectories max.
Example simulation re-computed with the new defaults.
Aggregation allowed for user-defined groups (argument my.location.file in pop.aggregate).
4.0-0 (5/20/2013)
-----
WPP data removed and put in separate packages (wpp2008, wpp2010). The package is now dependent on those.
=========
Odd version numbers refer to versions containing full (confidential) UN data.
Even version numbers refer to versions containing reduced (public) UN data.
===================================================
3.2-6 (4/19/2013)
-----
pop.map can processes expressions in parallel.
3.2-3/4 (4/10/2013)
-----
data are loaded into current environment.
3.2-1/2 (4/5/2013)
-----
.find.package replaced by find.package.
In expressions for M, Q and S, age groups 0-1, 1-4 and 0-4 are associated with indices -1, 0, 1, respectively.
3.2-0 (3/19/2012)
-----
Added mx, qx and migration to expressions.
Allowed using zero index for age group 0-1.
Functions LifeTableMx and LifeTableMxCol added.
Indicator added to pop.pyramid.
3.1-11 (2/16/2013)
------
Fixed an issue with NAs in TFR trajectories.
3.1-9/10 (1/3/2013)
-----
.First.lib and .Last.lib changed to .onLoad and .onUnload
3.1-7/8 (12/16/2012)
-----
Registred global object LOCATIONS (using utils::globalVariables()).
3.1-5/6 (12/15/2012)
-----
Example data reduced to two countries.
3.1-3/4 (12/02/2012)
-----
Bug fixes in regional aggregation.
Function pop.byage.plotAll added.
Suggested packages added to DESCRIPTION.
3.1-1 (11/16/2012)
-----
Bug in get.pop fixed.
In write.pop.projection.summary, argument 'alias' splitted into 'file.suffix' and expression.label.
Kannisto method modified to use female-male joint estimation of its slope parameter -> therefore no crossovers in the extrapolated death rates.
3.0-3 (10/14/2012)
-----
Male Mx for ages 100-130+ constrained to be larger than female Mx. Others no constraints.
3.0-1 (10/11/2012)
-----
Male Mx constrained to be larger than female Mx.
Functions for plotting and tabulating indicators by age (pop.byage.plot, pop.byage.table)
Functions for creating maps implemented (pop.map, pop.map.gvis).
Expression language added for visualizing and summarizing any population measure using pop.trajectories.plot/table and map functions.
Caching mechanism added for time-efficient plotting of maps.
Storing vital events possible via pop.predict (argument keep.vital.events).
Implemented a function for writing a summary file for any population measure and stored vital events (write.pop.projection.summary).
2.0-1/2 (06/05/2012)
-----
S3 object and S3 methods for plotting pyramids. The pyramid functions completly re-implemented.
Trajectories pyramid can now show any number of years.
An argument 'aggregation' added to get.pop.prediction.
1.0-1 (05/17/2012)
-----
Aggregation added (function pop.aggregate).
In addition to projection years, population pyramids can now show a past year.
Option for including another year into the population pyramids (argument draw.past.year).
0.2-1/2 (04/12/2012)
-----
Integrated outputs from joint female-male life expectancy.
+/- half child variant added.
0.2-0 (11/17/2011)
-----
Initial version delivered to the UN population division.