Collapse variable into groups and sum

This function makes it easy to collapse multiple labels of some column into groups, then sum them. The advantage of using this over simply relabeling a factor column (such as with forcats::fct_collapse) is that categories here don't have to be mutually exclusive. For example, from a table of populations by age group, you could collapse and aggregate into total population, people 18+, and people 65+ all within a single function call.

Usage

add_grps(data, grp_list, group = group, value = estimate, moe = NULL)

Arguments

data

A data frame; will honor grouping

grp_list

A named list of groups to collapse group into, either as characters that exactly match the labels in the grouping column, or as numbers giving the position of each label within unique values of the grouping column. Position numbers are easier to type correctly.

group

Bare column name giving groups in data; will be converted to factor

value

Bare column name of values. Defaults to estimate

moe

Bare column name of margins of error; if supplied, MOEs of sums will be included in output

Value

A data frame/tibble with sums of estimate. Retains grouping columns

Details

The quickest and most fool-proof way to create aggregate groups is to give their positions within a column's unique values. In this example column of ages:

1 ages 0-5
2 ages 6-17
3 ages 18-34
4 ages 35-64
5 ages 65-84
6 ages 85+

you would calculate total population from positions 1-6, ages 18+ from positions 3-6, and ages 65+ from positions 5-6. show_uniq() is a helper function for finding these positions.

See also

show_uniq()

Examples

# make a list of the positions of the groups you want to collapse
# e.g. education$edu_level[2:16] has the education levels that we consider
# less than high school
education |>
    dplyr::group_by(name) |>
    add_grps(
        list(
            ages25plus = 1,
            less_than_high_school = 2:16,
            high_school_plus = 17:25,
            bachelors_plus = 22:25
        ),
        group = edu_level, value = estimate
    )
#> # A tibble: 20 × 3
#> # Groups:   name [5]
#>    name       edu_level             estimate
#>    <chr>      <fct>                    <dbl>
#>  1 Bethany    ages25plus                3725
#>  2 Bethany    less_than_high_school      130
#>  3 Bethany    high_school_plus          3595
#>  4 Bethany    bachelors_plus            2193
#>  5 East Haven ages25plus               20768
#>  6 East Haven less_than_high_school     1724
#>  7 East Haven high_school_plus         19044
#>  8 East Haven bachelors_plus            5201
#>  9 Hamden     ages25plus               41017
#> 10 Hamden     less_than_high_school     2375
#> 11 Hamden     high_school_plus         38642
#> 12 Hamden     bachelors_plus           19132
#> 13 New Haven  ages25plus               84441
#> 14 New Haven  less_than_high_school    11853
#> 15 New Haven  high_school_plus         72588
#> 16 New Haven  bachelors_plus           31866
#> 17 West Haven ages25plus               35813
#> 18 West Haven less_than_high_school     4393
#> 19 West Haven high_school_plus         31420
#> 20 West Haven bachelors_plus           10399