Solve the base year of a marked horizon

Solves the base year on its own, as an own-problem variant, and returns the ledger of what it decided. This is the solve the projection does not contain: with base_year marked, the projection's LP has no base-year variables at all.

Usage

solve_base_year(
  object,
  name = NULL,
  solver = NULL,
  optimize_retirement = FALSE,
  variant = "by",
  save = TRUE,
  ...,
  verbose = TRUE
)

Arguments

object

a model, or an interpolated scenario, whose horizon marks a base_year.

name

scenario name; defaults to the object's own.

solver

solver options, as solve_scenario() takes.

optimize_retirement

allow the base year to retire stock endogenously.

variant

the variant label for the base-year run.

save

store the solved scenario, so a projection can refer to it by name later. TRUE by default: a base-year solution nothing can reopen cannot be carried into a scenario built in another session.

...

passed to interpolate_model() and solve_scenario().

verbose

logical.

Value

A list with scenario, ledger, horizon and base_year.

Details

Endogenous retirement is switched off for this solve unless optimize_retirement = TRUE. A past year's retirements are observed, not chosen, and a base year left free to retire its own fleet mis-dates every vintage carried forward.

What crosses to the projection is only what this solve decided — the ledger's newcap. The model's own @stock, which is where observed history lives and which already carries its vintages, is left alone.

The LP is the base year alone. Milestones before it are history and are solved here no more than they are in the projection; their data still anchors interpolation.

See also

solution_ledger(), newHorizon(), horizon_windows()

Examples

if (FALSE) { # \dontrun{
h <- newHorizon(intervals = iv, base_year = 2025, name = "h25")
mod <- setHorizon(mod, h)
by <- solve_base_year(mod, solver = solver_options$glpk)
by$ledger
} # }