mpascariu / mortalityestimate

Indirect Estimation Methods for Measurement of Demographic Indicators - R package

R package - Indirect Estimation Methods for Measurement of Demographic Indicators

---

This repository includes R code for estimating mortality indicators and full life tables based on given one or two pieces of information: life expectancy, child mortality, or child and adult mortality. The implemented models are better suited to the practical needs of mortality estimation, since the input parameters are continuous yet the second one is optional.

Installation

1. Make sure you have the most recent version of R
2. Run the following code in your R console

# install.packages("devtools")

library(devtools)
install_github("mpascariu/MortalityEstimate")

Help

All functions are documented in the standard way, which means that once you load the package using library(MortalityEstimate) you can just type ?LinearLink to see the help file.

Check the examples provided in the wilmoth and LinearLink functions.

Support

mortalityestimate's Issues

vx rotation procedure for theta > 0

A method for automatically estimating the parameters in the vx rotation procedure is needed.
So far the vx rotation can be applied only if theta = 0 (i.e. only if we reconstruct the entire mortality curve).

e(x) fit interpretation

Hi! great model, congrats. Simple and powerful.
I was exploring this with $\theta=80$, to focus in oldest ages. So I tried with the default example:
ages <- 0:100
years <- 1965:1990
sex <- 'female'
SWEmx <- HMD4mx$SWE[paste(ages), paste(years)]
M <- LinearLink(x = ages, mx = SWEmx, y = years, country = 'SWEDEN', theta = 80, method = 'LSE')
e80 <- 7
LT1 <- LinearLinkLT(M, ex = e80)
LT1$lt$ex[LT1$lt$x==80] #1.16

I was expecting something similar to 7 (when I play with $e_0$ prediction fits input). Am I missunderstanding something?
Thanks!

convertFx function

Dear pascariu
I have used the maximum entropy method to model the mortality of Iran, but unfortunately the convertFx function does not work well on my data, please help me

With warmest regards
Aida Aalabeygi

Confidence Intervals

Implement probabilistic confidence intervals for the reconstructed mortality curve.
I've tried a method so far, however because it is not very fast it's not included in the current version on the package.

Test the model for other levels of life expectancy

The current version of the package works very well with life expectancy at birth. The model needs to be adapted for remaining life expectancy any age. This means: making sure that all the objects inside the functions adapt to the different length of the input data. Coming asap.