Did the cartogram actually converge?

Cartograms fail quietly. An under-converged one looks entirely plausible while still misrepresenting the areas it exists to make honest. This reports the residual error per country, so the failure is visible.

Usage

cartogram_diagnostics(x, weight = NULL)

Arguments

x

A ggplot from cartogram_map() or dorling_map(), or the sf frame the cartogram was computed from.

weight

The weight column (unquoted). Required when x is a plain sf frame; read from the plot otherwise.

Value

A tibble of iso3c, target_share (the country's share of the weight), actual_share (its share of the cartogram's area) and area_error (the relative difference). The summary – mean absolute error, worst country – is attached as the "countryatlas_cartogram" attribute.

What counts as converged

A perfect cartogram has area_error of 0 everywhere. In practice a mean absolute error under a few percent is good and under 10% is usually acceptable; a systematically large error, or one concentrated in the small countries, means the algorithm stopped early. Raise itermax and try again.

See also

cartogram_map(), dorling_map(), gridded_cartogram()

Examples

# \donttest{
if (requireNamespace("sf", quietly = TRUE) &&
    requireNamespace("cartogram", quietly = TRUE) &&
    requireNamespace("rnaturalearth", quietly = TRUE)) {
  sfd <- attach_geometry(countryatlas::world_snapshot$countries,
                         geometry = "sf")
  cg <- cartogram_map(sfd, population)
  cartogram_diagnostics(cg)
}
#> # A tibble: 170 × 4
#>    iso3c target_share actual_share area_error
#>    <chr>        <dbl>        <dbl>      <dbl>
#>  1 GRL     0.00000702    0.00262      371.   
#>  2 BTN     0.0000978     0.000509       4.20 
#>  3 RUS     0.0177        0.0496         1.80 
#>  4 GUY     0.000103      0.000257       1.50 
#>  5 ISL     0.0000477     0.000101       1.11 
#>  6 NOR     0.000688      0.00144        1.10 
#>  7 NZL     0.000653      0.00133        1.03 
#>  8 CAN     0.00510       0.0103         1.03 
#>  9 LAO     0.000960      0.00177        0.844
#> 10 BHS     0.0000496     0.0000905      0.825
#> # ℹ 160 more rows
# }