In jschoeley/pash: Pace-Shape Analysis of Life-tables

options(width = 100)
knitr::opts_chunk$set(tidy = FALSE, size = "small")

What's it for?

pash lets you perform pace-shape analysis[^1][^2][^3] on life-tables. Calculate a wide array of pace and shape metrics and use them to standardize and compare your data.

pash is based on a powerful implementation of single-decrement life-tables. Allowing for various input formats, a choice of sensible defaults and extensive error-checking routines make the construction of life-tables extremely easy with pash while experts can choose among different life-table estimation methods.

Step 1: Create a pace-shape object from life table columns using Inputlx() or Inputmx().
Step 2: Using the pace-shape object as argument, calculate pace measures with GetPace(), calculate shape measures with GetShape(), standardize a life table by pace and shape with StandardizeLT(), re-base your life table to a different age with RebaseLT().

[^1]: Baudisch, Annette. 2011. "The Pace and Shape of Ageing." Methods in Ecology and Evolution 2 (4): 375–82. doi:10.1111/j.2041-210X.2010.00087.x. [^2]: Wrycza, Tomasz, and Annette Baudisch. 2014. "The Pace of Aging: Intrinsic Time Scales in Demography." Demographic Research 30 (1): 1571–90. doi:10.4054/DemRes.2014.30.57. [^3]: Wrycza, Tomasz F., Trifon I. Missov, and Annette Baudisch. 2015. "Quantifying the Shape of Aging."" PLOS ONE 10 (3): 1–18. doi:10.1371/journal.pone.0119163.

How to install?

You can install pash by running:

install.packages("devtools")
devtools::install_github("jschoeley/pash")

Examples

library(pash)

Create a life-table from mortality rates

Let's say we have a vector of mortality rates by age (Sweden, females, 1940--1944) and would like to calculate a life-table from it...

swe <- subset(sweden5x5, sex == "female" & period == "1940-1944")[c("x", "nmx")]
swe

...we use the Inputnmx() function and specify the last age group as open, done.

Inputnmx(x = swe$x, nmx = swe$nmx, last_open = TRUE)

We can change the details of the life-table calculation.

Inputnmx(x = swe$x, nmx = swe$nmx, last_open = TRUE, nax = "cfm")

Create a life-table from a survival curve

Given a vector of age-specific survival probabilities we can easily derive a whole life-table. By default the last age group is assumed to be closed.

prestons_lx
Inputlx(x = prestons_lx$x, lx = prestons_lx$lx)

Calculate measures of pace and shape

pash = Inputlx(x = prestons_lx$x, lx = prestons_lx$lx)

Once a pash object has been created using one of the Input*() functions pace and shape measures are readily available. By default all available pace measures are returned.

GetPace(pash)

The age where only 20% of the population is alive:

GetPace(pash, type = "qlx", q = 0.2)

By default all available shape measures are returned.

GetShape(pash)

The life-table entropy:

GetShape(pash, type = "entropy")

Rebase a life-table

Sometimes life-table shape and pace measures are only calculated for the adult ages. pash lets you "re-base" a life-table to a new age.

Create a pash object from a survival curve.

pash = Inputlx(x = prestons_lx$x, lx = prestons_lx$lx)
pash

Re-base to age 15.

pash = RebaseLT(pash, origin = 15)
pash

Calculate shape measures on the re-based life-table.

GetPace(pash)
GetShape(pash)

Re-basing is a non-destructive operation and thus can be reversed:

RebaseLT(pash, origin = 0)

Standardize a life-table by pace and shape

Different life-tables can be made comparable in shape by standardizing them by pace.[^3] pash allows you to standardize your life-table by different shape measures, total life expectancy being the default:

pash = Inputlx(x = prestons_lx$x, lx = prestons_lx$lx)
StandardizeLT(pash)

Standardize by median life-span:

StandardizeLT(pash, pace = "qlx")

Standardize by 0.7 quantile life-span:

StandardizeLT(pash, pace = "qlx", q = 0.7)

Using pash within dplyr pipe

Here we use pash in conjunction with dplyr and ggplot2 to produce a plot of life expectancy at birth versus life-table entropy for hundreds of Swedish life-tables.

library(tidyverse)

sweden5x5 %>%
  group_by(sex, period) %>%
    do({
     pash = Inputnmx(x = .$x, nmx = .$nmx)
     data.frame(e0 = GetPace(pash, type = "e0"),
                H  = GetShape(pash, type = "entropy"))
    }) %>%
  ggplot(aes(x = e0, y = H, color = sex)) + geom_point()