Normalize and classify per-capita nourishment.

Adds a piecewise-normalized nourishment score value_norm and its Under / Adequate / Over class nourish. The normalization (the Global Intake_normalization) is value / floor below the floor, 1 + value / ceiling above the ceiling and 1 + (value - floor) / (ceiling - floor) in between, so the score is below 1 under the floor, exactly 1 at the floor, between 1 and 2 across the adequate band and 2 or above at or past the ceiling. The class is "Under" when value_norm < 1, "Adequate" when value_norm < 2 and "Over" otherwise. Protein is the SJOS-N nourishment axis, so the defaults are the protein floor and ceiling (62.1 and 85.05 g/cap/day) from nourishment_thresholds.

Of those two defaults only the underlying 46 g/cap/day floor is sourced (WHO/FAO/UNU TRS 935 Table 46, the safe intake of a 55 kg adult, itself a 97.5th-percentile individual level rather than a population one). The 63 ceiling and the 1.35 factor that lifts both to a supply basis carry no source; nourishment_thresholds$provenance says so per row. build_nourishment_band() is the sourced replacement for both bounds and is not wired in here yet.

Passing value_col = energy_kcal_cap_day runs the same arithmetic on a different quantity and is not a second WHEP axis: the packaged energy bounds are unsourced, and WHEP's energy column is gross combustion energy where a dietary threshold is metabolisable. Supply your own bounds and your own metabolisable series if you want that comparison.

Usage

normalize_nourishment(x, value_col = protein_g_cap_day, thresholds = NULL)

Arguments

x

A tibble carrying the per-capita nourishment column named by value_col (for example a build_food_supply() output).

value_col

The unquoted nourishment column to normalize. Defaults to protein_g_cap_day.

thresholds

Either a named floor/ceiling pair applied to every row (a named numeric vector or list), or a data frame of per-country-year bounds keyed by year and area_code with either floor_g_cap_day/ceiling_g_cap_day or floor/ceiling — so a build_nourishment_band() output passes straight through. A row that matches no band is classified NA and named in a warning, never silently given the flat default. When NULL (default) the flat protein bounds from nourishment_thresholds are used.

Value

x with value_norm (numeric score) and nourish ("Under", "Adequate" or "Over") added.

Examples

normalize_nourishment(
  tibble::tribble(
    ~area_code, ~protein_g_cap_day,
    10L, 30,
    20L, 70,
    30L, 100
  )
)
#> # A tibble: 3 × 4
#>   area_code protein_g_cap_day value_norm nourish 
#>       <int>             <dbl>      <dbl> <chr>   
#> 1        10                30      0.483 Under   
#> 2        20                70      1.34  Adequate
#> 3        30               100      2.18  Over