This file details how the inputs files for the wedding grid model were created.


R was used to create these files. R is a free and open source statistical software package and is available from http://cran.r-project.org/.  

R studio, an IDE for R (available here: http://www.rstudio.com/) was also used, although this is not essential.


These inputs involve a combination of empirical data and forecasts. Packaged with this readme are the original empirical data and the R scripts necessary to conduct the forecasting. 
The R Scripts have explanatory annotations, but please contact the authors with any questions. 

You may need to install further R packages to run the script. In particular, the package "forecast" is a prerequisite (see the reference in the Bibliography of the accompanying paper). 
This is also available from http://cran.r-project.org/ , and more information is available from http://robjhyndman.com/software/forecast/ . 




Four input files are needed by the wedding grid:

These are included in the version of the simulation available on the OpenABM site, but are included here so that the details of the forecasting are made explicit.



1. Death.rate.new.female.csv

2. Death.rate.new.male.csv

3. Fertility.rate.new.csv

4. birthprops.csv

Files 1 and 2 are created using the empirical data from "Exposures_1x1.csv", "Deaths_1x1.csv" and the R scripts "mortality_Script.R " and "mortality_functions.R "  
The empirical data is for the UK and taken from the Human Mortality Database, and re-saved as csv files. (Data Link http://www.mortality.org/cgi-bin/hmd/country.php?cntr=GBR&level=2 ) 


Unzip the files to one folder and change the first line of code in mortality_script.R so that it sets the working directory to the location you unzipped to. ( In R studio you can use the " tools> set working directory > to source file location " option ) 
Next, you should be able to run entireity of the R script, and the output files 1 and 2  will be created. Some plots of the forecasts will also be displayed.
 

File 3 is created from the empirical data from the "fertdata.csv" and "cohortasfr.csv" files. This data is taken from Eurostat and from an ONS fertility dataset respectively. 
(Data Links: http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=demo_frate&lang=en, and table 10.1 at 
http://www.ons.gov.uk/ons/rel/vsob1/birth-statistics--england-and-wales--series-fm1-/no--27--1998/birth-statistics-series-fm1.pdf) 
The R scripts "fertilty_script.R" and "fertility_functions.R" are needed to create the input file.

As with files 1 and 2, running the R script once the working directory has been changed will create the fertility.rate.new.csv file and also display a few plots. 

File 4 is created from the  "fertsplit.csv" data file and the "birthpropcalcs.R" script. 
(Data Link http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=demo_fagec&lang=en)
 The fertsplit.csv file gives the proportion of children to married mothers by age, where that data was available from Eurostat. Again, just change the working directory and run.




Please contact jdh4g10@soton.ac.uk with any problems or questions. 


Also please examine the excellent demography package by the same author as the forecast package, and which was used a point of reference for the implementation of some of the functions used in scripts above.
 Reference: 
Hyndman, R. (2011). demography: Forecasting Mortality and Fertility Data. R Package. Accessed on 
26/07/2011.
URL http://robjhyndman.com/software/demography/