
Expand grouped population to single-year ages while preserving an open group
Source:R/age.R
expand_age_pop.Rdexpand_age_pop() converts any set of closed age groups to single-year
estimates and keeps the final source group as an actual open age group. It
never represents an open group as a fabricated sequence of high ages. Every
effective source-group total is constrained to its input total exactly, and
estimates are non-negative.
Arguments
- x
Numeric population totals. Its length must equal
length(source_breaks); the last value is the open age group.- source_breaks
Integer lower bounds of the input age groups. The last bound starts the open group. The default is 0, 1–4, 5–9, ..., 85+.
- method
Splitting method:
"linear","constant","fmm","natural","monoh.fc","reference","sprague", or"pclm". The spline options are retained for compatibility and apply only to closed groups."reference"uses an external single-age schedule."sprague"requires regular closed five-year groups; the default 0 and 1–4 groups are automatically combined to 0–4."pclm"requiresungroup.- reference
An external single-age schedule for
method = "reference". Supply a named numeric vector (names are ages), or a data frame withage/xandpopulation/pop/ycolumns. It must cover every closed age to be split. It should be independent of data evaluated in a blind validation.- high_age
A character method or list supplied to
expand_open_age_pop(). The default,"keep", retains the source open group. Use a list withmethod = "reference",reference, andmax_ageonly when an external high-age schedule is available.- pclm_control
A list passed to
ungroup::pclm(control = ...).
Value
A data frame with x (age lower bound), y (population), age_end,
open, source_group, lower, and upper. Closed rows have
age_end == x; the final open row has age_end = Inf. PCLM uncertainty
intervals are returned when supplied by ungroup; other methods report
missing intervals. Attributes record the effective source_breaks, method,
and uncertainty method.
References
Sprague, T. B. (1880). Explanation of a new formula for interpolation. Journal of the Institute of Actuaries, 22, 270–285.
Rizzi, S., Gampe, J., and Eilers, P. H. C. (2015). Efficient ungrouping of coarse histograms and age heaping in demographic data. Statistical Methods in Medical Research, 24(3), 291–313.
Examples
population <- c(
5053, 17743, 25541, 32509, 30530, 34806, 36846, 38691, 40056,
39252, 37349, 30507, 26363, 21684, 15362, 11725, 7461, 3260, 915
)
expand_age_pop(population, method = "sprague")
#> x y age_end open source_group lower upper
#> 1 0 4985 0 FALSE 1 NA NA
#> 2 1 4622 1 FALSE 1 NA NA
#> 3 2 4420 2 FALSE 1 NA NA
#> 4 3 4357 3 FALSE 1 NA NA
#> 5 4 4412 4 FALSE 1 NA NA
#> 6 5 4564 5 FALSE 2 NA NA
#> 7 6 4792 6 FALSE 2 NA NA
#> 8 7 5074 7 FALSE 2 NA NA
#> 9 8 5391 8 FALSE 2 NA NA
#> 10 9 5720 9 FALSE 2 NA NA
#> 11 10 6087 10 FALSE 3 NA NA
#> 12 11 6514 11 FALSE 3 NA NA
#> 13 12 6755 12 FALSE 3 NA NA
#> 14 13 6697 13 FALSE 3 NA NA
#> 15 14 6456 14 FALSE 3 NA NA
#> 16 15 6245 15 FALSE 4 NA NA
#> 17 16 6005 16 FALSE 4 NA NA
#> 18 17 5905 17 FALSE 4 NA NA
#> 19 18 6043 18 FALSE 4 NA NA
#> 20 19 6332 19 FALSE 4 NA NA
#> 21 20 6585 20 FALSE 5 NA NA
#> 22 21 6843 21 FALSE 5 NA NA
#> 23 22 7047 22 FALSE 5 NA NA
#> 24 23 7148 23 FALSE 5 NA NA
#> 25 24 7183 24 FALSE 5 NA NA
#> 26 25 7238 25 FALSE 6 NA NA
#> 27 26 7294 26 FALSE 6 NA NA
#> 28 27 7356 27 FALSE 6 NA NA
#> 29 28 7435 28 FALSE 6 NA NA
#> 30 29 7523 29 FALSE 6 NA NA
#> 31 30 7599 30 FALSE 7 NA NA
#> 32 31 7665 31 FALSE 7 NA NA
#> 33 32 7733 32 FALSE 7 NA NA
#> 34 33 7809 33 FALSE 7 NA NA
#> 35 34 7885 34 FALSE 7 NA NA
#> 36 35 7952 35 FALSE 8 NA NA
#> 37 36 8013 36 FALSE 8 NA NA
#> 38 37 8046 37 FALSE 8 NA NA
#> 39 38 8040 38 FALSE 8 NA NA
#> 40 39 8005 39 FALSE 8 NA NA
#> 41 40 7957 40 FALSE 9 NA NA
#> 42 41 7886 41 FALSE 9 NA NA
#> 43 42 7828 42 FALSE 9 NA NA
#> 44 43 7800 43 FALSE 9 NA NA
#> 45 44 7781 44 FALSE 9 NA NA
#> 46 45 7743 45 FALSE 10 NA NA
#> 47 46 7707 46 FALSE 10 NA NA
#> 48 47 7583 47 FALSE 10 NA NA
#> 49 48 7327 48 FALSE 10 NA NA
#> 50 49 6989 49 FALSE 10 NA NA
#> 51 50 6655 50 FALSE 11 NA NA
#> 52 51 6302 51 FALSE 11 NA NA
#> 53 52 6010 52 FALSE 11 NA NA
#> 54 53 5827 53 FALSE 11 NA NA
#> 55 54 5713 54 FALSE 11 NA NA
#> 56 55 5578 55 FALSE 12 NA NA
#> 57 56 5439 56 FALSE 12 NA NA
#> 58 57 5290 57 FALSE 12 NA NA
#> 59 58 5121 58 FALSE 12 NA NA
#> 60 59 4935 59 FALSE 12 NA NA
#> 61 60 4756 60 FALSE 13 NA NA
#> 62 61 4589 61 FALSE 13 NA NA
#> 63 62 4385 62 FALSE 13 NA NA
#> 64 63 4124 63 FALSE 13 NA NA
#> 65 64 3830 64 FALSE 13 NA NA
#> 66 65 3542 65 FALSE 14 NA NA
#> 67 66 3247 66 FALSE 14 NA NA
#> 68 67 3002 67 FALSE 14 NA NA
#> 69 68 2840 68 FALSE 14 NA NA
#> 70 69 2731 69 FALSE 14 NA NA
#> 71 70 2615 70 FALSE 15 NA NA
#> 72 71 2505 71 FALSE 15 NA NA
#> 73 72 2376 72 FALSE 15 NA NA
#> 74 73 2209 73 FALSE 15 NA NA
#> 75 74 2020 74 FALSE 15 NA NA
#> 76 75 1841 75 FALSE 16 NA NA
#> 77 76 1667 76 FALSE 16 NA NA
#> 78 77 1492 77 FALSE 16 NA NA
#> 79 78 1317 78 FALSE 16 NA NA
#> 80 79 1144 79 FALSE 16 NA NA
#> 81 80 974 80 FALSE 17 NA NA
#> 82 81 808 81 FALSE 17 NA NA
#> 83 82 646 82 FALSE 17 NA NA
#> 84 83 490 83 FALSE 17 NA NA
#> 85 84 342 84 FALSE 17 NA NA
#> 86 85 915 Inf TRUE 18 NA NA