Skip to contents

Plot rolling-mean year-on-year drivers of the change in territorial N losses

Source: R/decomposition_analysis.R
plot_loss_decomp_rolling.Rd

Same data as plot_loss_decomp_yearly(), but smooths each year's own additive contribution with a centered rolling mean (window years wide, NA-padded at the edges) before plotting, to make sustained multi-year trends (e.g. a period of continuously improving efficiency) easier to see than in the raw, noisy year-on-year series.

Usage

plot_loss_decomp_rolling(decomp = NULL, window = 10)

Arguments

decomp

A named list from decompose_terr_losses(). If NULL, computed automatically (slow).

window

Width of the centered rolling-mean window, in years. Default 10.

Value

A named list with ggplot objects by_compartment and by_mechanism.

Examples

# `window` must not exceed the number of years available per group.
decomp <- list(
  by_compartment = tibble::tribble(
    ~t0, ~compartment, ~contribution_mgn, ~cumulative_mgn,
    1861, "cropland", 12000, 12000,
    1862, "cropland", 8000, 20000,
    1863, "cropland", 9000, 29000
  ),
  by_mechanism = tibble::tribble(
    ~t0, ~mechanism, ~contribution_mgn, ~cumulative_mgn,
    1861, "Size", 9000, 9000,
    1862, "Size", 5000, 14000,
    1863, "Size", 6000, 20000
  )
)
plots <- plot_loss_decomp_rolling(decomp, window = 3)