Reproducable example of projection using mstls

example_riffe.r

library(devtools)
dev_mode(on=TRUE)
install_github(username='mhermans', repo='mstls')

library(mstls)

data(YU61female) # included dataset in mstls: 2-region female population (Yugoslavia, 1961)

# calculate observed rates
obs_rates <- with(
  YU61female, 
  cbind(
    to_SL/population,
    to_RYU/population,
    deaths/population,
    births/population
  )
)

obs_rates <- cbind(
  rep(seq(0, 85, 5), 2), 
  rep(seq(4, 89, 5), 2), 
  rep(c(1,2), each=18), 
  obs_rates)

colnames(obs_rates) <- c('x', 'xn', 'state', 'Mx_i1', 'Mx_i2', 'Mdx', 'Fx')

Mx.ex <- transfer_matrix(obs_rates[,-7], multiple=TRUE) # drop Fx
Px.ex <- transfer_prob(Mx.ex, multiple=TRUE)
lx.ex <- expected_survivors(Px.ex, 100000)
Lx.ex <- years_lived(lx.ex, Mx.ex)  
Sx.ex <- survivor_prop(years_lived=Lx.ex)
Bx.ex <- birth_prop(birth_rates=obs_rates[,7], transition_probs=Px.ex, survivor_prop=Sx.ex)
G.ex <- projection_matrix(birth_prop=Bx.ex, survivor_prop=Sx.ex)

dim(G.ex)

# add labels to G
a <- rep(seq(0,80,5), each=2)
b <- rep(1:2, 17)

rlbls <- c(c('Bx_i1', 'Bx_i2'), paste('S', a, '_i', b, sep=''))

a <- rep(seq(0,85,5), each=2)
b <- rep(1:2, 18)

clbls <- paste(a, '_', b, 'j', sep='') 

dimnames(G.ex) <- list(rlbls, clbls)
rm(a, b, clbls, rlbls)

# project 8 steps
N.projected <- project(YU61female$population, n_states=2, pmat=G.ex, n_steps=8)
rownames(N.projected) <- rep(seq(0, 85, 5), 2)
N.projected