Description Usage Arguments Details Value Author(s) References See Also Examples

Creates a list of observed and predicted period life tables across from YourCast output for each geographical unit. This function should only be used when forecasting all-cause mortality rates.

1 |

`obj` |
A |

`ax` |
A scalar, numeric vector of length |

`a0` |
A scalar indicating the average years lived within the first age interval by individuals who die within that age interval. Ignored if |

`nx` |
A scalar or vector of length |

`dv.log` |
Logical. Is the dependent variable logged? If |

The input must be a `yourcast`

object containing forecasts of all-cause mortality rates. The death rates do not have to be for one-year age intervals, albeit `ax`

and `a0`

should be adjusted accordingly based on the age interval size. The rates may either be logged or unlogged. If `dv.log`

is set to `NULL`

, the function parses the formula automatically to determine whether the rates are logged. Setting `dv.log`

to `TRUE`

or `FALSE`

manually sets whether the rates are logged or not.

Specifying `ax`

is necessary to convert mortality rates into conditional probabilities of death. Suppose that *m_x* is the mortality rate for the *n*-year age interval *[x,x+n)*. *a_x* is then defined as the average years lived within the age interval *[x,x+n)* by individuals who die at that age. The default value of *a_x* is 0.5, which means that individuals on average live half the year within the age that they die (for single-year age intervals). The default value of *a_0* is set to 0.07, following the reported values within the Human Mortality Database for adult females in the United States for the *[0,1)* age interval in the last decade.

Alternatively, one can specify `ax`

as a vector of length *A*, where *A* is the number of age intervals. In this case, *a_x* is assumed to be constant across time for each age interval. One can also pass a *T* by *A* matrix, in which case *a_x* is allowed to vary over time.

The last age interval is open-ended: *[x,∞)*. The conditional probability of death is set to 1 and *a_x* is automatically set to *1/m_x*.

If a single geographic area, a list with names `obs`

and `pred`

is returned, where the elements are life table arrays calculated using observed and predicted mortality rates, respectively. If multiple geographic areas, a named list of geographic areas is returned, where each element contains a list of observed and predicted life tables.

Each life table array has dimensions *T* by *A* by *10*, where the third dimension contains the following variables:

`x` |
Age interval. |

`nx` |
Length of age interval. |

`ax` |
Average years lived within age interval by individuals who die during that age interval. |

`qx` |
Conditional probability of death for the age interval. |

`px` |
Conditional probability of survival for the age interval. |

`lx` |
Number of survivors at the start of the age interval. Set to 100,000 for the first age interval. |

`dx` |
Number of individuals dying within the age interval. |

`Lx` |
Person-years lived by individuals in the age interval. |

`Tx` |
Person-years remaining at start of age interval. |

`ex` |
Remaining life expectancy at start of age interval. |

Konstantin Kashin [email protected]

Preston, S.H., Patrick Heuveline, Michel Guillot. (2000) *Demography: Measuring and Modeling Population Processes.* Blackwell Publishing.

Wilmoth, J.R. et al. “Methods Protocol for the Human Mortality Database” http://www.mortality.org/.

http://gking.harvard.edu/yourcast

`yourcast`

function and documentation
(`help(yourcast)`

)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | ```
# Run Lee-Carter model for Figure 2.6 in Demographic Forecasting
data(chp.2.6.1)
ff.allc <- log(allc2/popu2) ~ time
ylc.allc <- yourcast(formula=ff.allc, dataobj=chp.2.6.1, model="LC",
elim.collinear=FALSE,
sample.frame=c(1950,2000,2001,2060))
# calculate lifetable function assuming ax=2.5 and a0=1.907
# a0 calculated for both males and females from
# Human Mortality Database for New Zealand for 2000
lt <- lifetable(ylc.allc,ax=2.5,a0=1.907)
# observed lifetable for csid 5150 (New Zealand) for 2000
lt[["5150"]]$obs["2000",,]
# predicted lifetable for csid 5150 (New Zealand) for 2015
lt[["5150"]]$pred["2015",,]
``` |

IQSS/YourCast documentation built on May 8, 2017, 8:03 p.m.

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.