Extract a tidy mix (generation, capacity, fuel) from a solved scenario

Collects the building blocks of a "mix" chart from a solved scenario into one tidy data.frame: technology output, storage charge/discharge, inter-regional and rest-of-world trade, and demand. The same extractor feeds autoplot(scenario, ...), the scenario report, and any custom analysis (including sonification – see the useR! training course).

Usage

getMix(
  scen,
  type = c("generation", "capacity", "new_capacity", "fuel"),
  comm = "ELC",
  region = NULL,
  year = NULL,
  timeslice = NULL,
  drop_small = 0,
  by = NULL,
  top_n = NULL
)

Arguments

scen

a solved scenario object, or a named list of scenarios (results are row-bound with the list names in the scenario column).

type

character: "generation" (default – output of comm by process, with storage-in and exports negative), "capacity" (vTechCap + vStorageOutCap), "new_capacity" (vTechNewCap + vStorageOutNewCap), or "fuel" (technology fuel consumption vTechInp by input commodity).

comm

character, the balanced commodity for "generation" (default "ELC"). Ignored for the other types.

region

character vector or NULL (all regions). Inter-regional trade flows are only meaningful for a region subset and are skipped when region = NULL (they cancel out in an all-region sum).

year

integer vector or NULL (all milestone year).

timeslice

NULL for annual sums, or a regular expression selecting a timeslice sample (e.g. "^SUM_" for the summer day on the s4_h24 calendar). When the matched timeslices carry an hour tag ("_h00"..."_h23"), an integer hour column is added.

drop_small

numeric in [0, 1): drop processes whose total absolute value is below this share of the largest process (default 0, keep all). Unlike top_n this DELETES rows – stacked charts lose that mass.

by

character vector, any of "vintage" and "cluster". Technology variants are always rolled back up to their base technology in process; by default (NULL) they are simply summed, so charts do not fragment into one series per variant. Naming a dimension here keeps it as an extra grouping column (with "(none)" for processes that have no variants), ready to drive a fill or facet.

top_n

integer: keep the top_n largest processes (by summed absolute value) and lump the rest into one "Other" series. Mass-preserving – lumped rows are re-aggregated, not dropped – and the demand overlay is never lumped. NULL (default) or Inf keeps every process. Applied after drop_small. On a list of scenarios the lumping runs per scenario, so each keeps its own top set.

Value

A tidy data.frame with columns scenario, type, process, flow (generation, storage-in/out, import/export, demand, fuel, capacity, new_capacity), comm, region, year, value, and – when timeslice is given – timeslice (+ hour when parsable), plus any column named in by. Missing variables are skipped silently (e.g. a model without storage or trade).

Examples

if (FALSE) { # \dontrun{
gen <- getMix(scen, "generation")                      # annual, all regions
day <- getMix(scen, "generation", timeslice = "^SUM_")     # summer-day dispatch
cmp <- getMix(list(BASE = s1, CO2CAP = s2), "capacity")
cl  <- getMix(scen, "capacity", by = "cluster")        # split by cluster
} # }