life {cNORM} | R Documentation |
The data is available by the courtesy of the World Bank under Creative Commons Attribution 4.0 (CC-BY 4.0). It includes the life expectancy at birth on nation level from 1960 to 2017. The data has been converted to long data format, aggregates for groups of nations and missings have been deleted and a grouping variable with a broader scope spanning 4 years each has been added. It shows, that it can be better to reduce predictors. The model does not converge anymore after using 8 predictors and the optimal solution is achieved with four predictors, equaling R2=.9825.
life
A data frame with 11182 rows and 4 variables:
The name of the country
reference year of data collection
the life expectancy at birth
a grouping variable based on 'year' but with a lower resolution; spans intervals of 4 years each
https://data.worldbank.org/indicator/sp.dyn.le00.in
The World Bank (2018). Life expectancy at birth, total (years). Data Source World Development Indicators available https://data.worldbank.org/indicator/sp.dyn.le00.in (date of retrieval: 01/09/2018)
## Not run: # data preparation data.life <- rankByGroup(life, raw="life") data.life <- computePowers(data.life, age="year") #determining best suiting model by plotting series model.life <- bestModel(data.life, raw="life") plotPercentileSeries(data.life, model.life, end=10) # model with four predictors seems to work best model2.life <- bestModel(data.life, raw="life", terms=4) ## End(Not run)