Package 'CRIO'

Description

GFSir parses the 4th id segment into co2, so B4T suffixes land there. Everything is derived from id, ssp, gcm and rcp instead.

Usage

add_scenario_cols(
  df,
  pathways = default_pathways,
  low_warming_gcms = c("GFDL", "MPI", "MRI"),
  lower_emission_rcps = "2.6"
)

Arguments

Details

Adds:

b4t

"None", the CG region where B4T is implemented, or "All CG regions".

ensemble

"Climate" or "B4T".

pathway

Name of the SSP x RCP combination from pathways (e.g. "Green world"), or "SSPx-RCPy" if not listed.

warming

"Lower emission" for lower_emission_rcps; otherwise "Low warming" / "High warming" by GCM; "No climate change" for NoCC.

and resets co2 / co2_fert (all runs are without CO2 fertilization).

Value

df with the added columns.

Description

Aquatic commodities are missing from Nutrients_per_kg.gdx. These are rough, globally generic approximations (same value for every country), informed by FAO/INFOODS uFiSh ranges and USDA food composition data for representative species in each IMPACT aquatic category.

Usage

aquatic_nutrient_per_kg

Format

A tibble with columns cmdty, Nutrients and value.

Details

Caveats: not country-specific; rough averages across highly variable species groups (small whole fish vs. large fillet fish differ by an order of magnitude for vitamin A, iron, zinc, folate). Treat as a placeholder and flag as a known limitation. Units match Nutrients codes, per kg edible portion, raw.

Description

The six CG regions plus "All CG regions" (the CG6 scenario).

Usage

b4t_levels

Format

A character vector.

Description

CGIAR regions where Breeding For Tomorrow (B4T) is implemented

Usage

b4t_regions

Format

A character vector of CG region codes.

Description

Each B4T run against the climate run with the same SSP, GCM and RCP (climate_counterpart(), e.g. SSP3-MRI-370-WCA vs SSP3-MRI-370-379) in target_year, so B4T runs under any scenario are handled. gap = distance from target (0 if met); gap_closed = share of the counterfactual's gap closed by B4T (NA if it already meets the target).

Usage

build_b4t_comparison(
  df,
  value_col,
  threshold = NULL,
  threshold_col = NULL,
  direction = c("below", "above"),
  target_year = 2050
)

Arguments

Value

One row per region x B4T run with counterfactual (base_id, ⁠base_*⁠) and B4T values, absolute and percent change, gaps, gap closed and target status change.

Description

B4T comparisons for a list of indicator specs

Usage

build_b4t_comparisons(specs, ...)

Arguments

Value

All comparisons stacked, with an indicator column.

Description

For each region, the share of scenarios hitting the target in target_year, and a classification tree (rpart) on the scenario descriptors for regions where some but not all scenarios hit it. Run on the climate ensemble (filter_climate()).

Usage

build_pathway_analysis(
  df,
  value_col,
  threshold = NULL,
  threshold_col = NULL,
  direction = c("below", "above"),
  target_year = 2050,
  predictors = c("pathway", "warming")
)

Arguments

Value

A list with df_year, summary_tbl, partial_regions, tree_results and rules_tbl.

Description

Read from the regions sheet of a GFSir mapping file, which is long: one row per (cty, region) membership. Cached after the first call.

Usage

cg_lookup(mapping = "mappingCG6.xlsx")

Arguments

Value

A tibble with columns cty and cg_region.

Description

The climate (non-B4T) run with the same SSP, GCM and RCP, e.g. SSP3-MRI-370-WCA -> SSP3-MRI-370-379. Used to compare each B4T run with its own counterfactual.

Usage

climate_counterpart(id)

Arguments

Value

Scenario id(s) of the matching climate run.

Description

SSP3, MRI climate model, RCP 7.0, no CO2 fertilization.

Usage

crio_baseline_id

Format

A character string.

Description

Runs every figure type for a set of indicators and writes them to plotdir. Each figure type and each indicator can be switched on or off, and a failing figure is reported and skipped instead of stopping the run.

Usage

crio_plot_all(
  fulldata,
  plotdir,
  plots = c("maps", "pct_hits", "rules", "trajectories", "trajectories_countries",
    "quantities", "overview", "b4t_heatmap", "b4t_gap"),
  indicators = NULL,
  nutrients = TRUE,
  regions = c(b4t_regions, "WLD"),
  countries = NULL,
  ctymap = NULL,
  specs = NULL,
  nut_threshold = 1.5,
  water_threshold = 1,
  keep_above = -0.1,
  b4t_map_scenario = "All CG regions",
  b4t_show_regions = c("all", "own"),
  b4t_show_countries = c("own", "all"),
  baseline_id = crio_baseline_id,
  target_year = 2050,
  subdirs = TRUE,
  format = "png"
)

Arguments

Details

Figure types (plots):

maps

4-panel indicator maps (climate ensemble). Needs ctymap.

pct_hits

Bar chart of percent of climate scenarios hitting the target.

rules

Heatmap of decision-tree rules across regions.

trajectories

Trajectories for regions (all B4T runs).

trajectories_countries

Trajectories for countries (own-region B4T).

quantities

Nutrient quantity trajectories (RNI and target lines), for regions and, if given, countries.

overview

Distance-to-target overviews (main indicators and nutrients), for regions and, if given, countries.

b4t_heatmap

Region x B4T heatmaps (absolute and percent change).

b4t_gap

Gap closed by B4T (bars) and gap maps for b4t_map_scenario.

Value

Invisibly, a list with analyses (pathway analyses by indicator), b4t (all B4T comparisons) and failed (figures that errored, with the message).

Description

Combinations not listed here are named automatically as "SSPx-RCPy". Pass your own table to add_scenario_cols() (via load_crio_data()) to rename or add pathways.

Usage

default_pathways

Format

A tibble with columns ssp, rcp (as parsed by GFSir, e.g. "2.6", "7.0") and pathway.

Description

Raw/intermediate commodities that would double count, plus palm kernel (food availability only in IDN and MYS, and small).

Usage

exclude_cmdty

Format

A character vector of IMPACT demand commodity codes.

Description

Extract rules from a classification tree

Usage

extract_rules(tree, region_name)

Arguments

Value

A data frame with region, outcome ("hits"/"misses"), rule, cover.

Description

Fraction of a commodity's calories that come from fat (not per 100 g). Used to approximate the fat share of dietary energy.

Usage

fat_kcal_share

Format

A tibble with columns name, fat_kcal_share and notes.

Description

Drops B4T runs, leaving the SSP x GCM x RCP runs (including the baseline).

Usage

filter_climate(df)

Arguments

Value

Filtered df.

Description

Food groups used for calorie and water totals

Usage

food_groups

Format

A character vector of GFSir food group codes.

Description

ADER comes from FAOSTAT Food Security and is extended to future years with a linear trend per country.

Usage

get_ader(calories, faostat_csv, sets_xlsx, years = 2025:2050)

Arguments

Value

Total calories with ader and hits_target.

Description

Adds a Total row over food_groups and each row's share of the total. GRL is dropped (no demand except fish).

Usage

get_calories(files, mapping = "mappingCG6.xlsx")

Arguments

Value

Calories (kcal/person/day).

Description

Approximate share of calories from fat

Usage

get_fat_share(calories, fat_shares = fat_kcal_share)

Arguments

Value

One row per region x year x scenario with fat_energy_share.

Description

Food availability (all commodities and food groups)

Usage

get_foodavail(files, mapping = "mappingCG6.xlsx")

Arguments

Value

GFSir foodavail output for all scenarios (no scenario columns).

Description

Fruit and vegetable availability (g/person/day)

Usage

get_frtveg(foodavail)

Arguments

Value

F&V availability.

Description

Hunger is only solved in 5-year steps; GFSir's group5 merges it onto annual population, so the in-between years come back as 0. Only years with hunger results are kept.

Usage

get_hunger(files, mapping = "mappingCG6.xlsx")

Arguments

Details

Note: the hunger module floors the share at risk at 1 percent, scaled to each country's FAO 2020 numbers, so countries can flatten at a country-specific minimum (e.g. ~3.3 percent for CIV).

Value

Hunger data with value (million people) and share.

Description

Total calories against the minimum dietary energy requirement (MDER)

Usage

get_mder(calories, minkcal)

Arguments

Value

Total calories with minKCAL and hits_target.

Description

Minimum calorie requirement by country

Usage

get_minkcal(gdx)

Arguments

Value

A data frame with region and minKCAL.

Description

Food availability (kg/person/day) x nutrient content per kg, summed over commodities, divided by the RNI. Nutrient contents come from the Hunger model's Nutrients_per_kg.gdx (country-specific) plus aquatic_nutrient_per_kg placeholders for aquatic foods.

Usage

get_nutrient_adequacy(
  foodavail,
  nutrients_gdx,
  mapping = "mappingCG6.xlsx",
  rni = rni_lookup,
  aquatic = aquatic_nutrient_per_kg,
  exclude = exclude_cmdty
)

Arguments

Details

This is nutrient supply, not intake.

Value

One row per region x year x scenario x nutrient with total_nutrient (per capita per day), RNI and adequacy_ratio.

Description

Share of calories from sugar

Usage

get_sugar_share(calories, sugar_name = "O&S-Sugar  (p)")

Arguments

Value

One row per region x year x scenario with sugar_energy_share.

Description

Countries with total blue water below min_volume in the baseline in 2025 are dropped. ratio2025 is each value over its own 2025 value.

Usage

get_water(
  files,
  mapping = "mappingCG6.xlsx",
  baseline_id = crio_baseline_id,
  min_volume = 0.01
)

Arguments

Value

Blue water data (cubic km).

Description

One entry per indicator with its data, target and labels. Used to drive trajectory figures, distance-to-target overviews and B4T comparisons in a loop.

Usage

indicator_specs(fulldata, nut_threshold = 1.5, water_threshold = 1)

Arguments

Details

Each entry has: df, value_col, threshold or threshold_col, direction ("below" or "above" the target is good), label, optional short_label, labeller (axis labels) and delta_labeller (absolute change vs baseline; shares in percentage points).

Value

A named list of specs. The names of the non-nutrient indicators are stored in attr(, "main").

Description

Climate runs are named SSP-GCM-RCP-379 (no CO2 fertilization), e.g. SSP1-GFDL-126-379 or SSP3-MRI-370-379. B4T runs replace the last segment with the CG region where B4T is implemented (CWANA, ESA, ..., CG6 = all CG regions). Any SSP (1-5), GCM and RCP is picked up; you can also pass your own vector of files to the readers instead.

Usage

list_scenarios(scendir, include_nocc = FALSE)

Arguments

Value

Character vector of full file paths.

Description

Build all RIO indicator datasets

Usage

load_crio_data(
  files,
  nutrients_gdx,
  faostat_csv,
  sets_xlsx,
  mapping = "mappingCG6.xlsx",
  baseline_id = crio_baseline_id
)

Arguments

Value

A named list: hunger, fruits_and_vegetables, calories, fats, sugar, MDER, ADER, RNI, WATER, minkcal.

Description

Load the IMPACT region shapefile

Usage

load_region_map(shp)

Arguments

Value

A SpatVector.

Description

Display names for nutrient codes

Usage

nutrient_labels

Format

A named character vector, names are Nutrients codes.

Description

Nutrient availability in physical units (per capita per day) instead of adequacy ratios. The target is nut_threshold x RNI; the RNI itself is added as a reference line (ref_value).

Usage

nutrient_quantity_specs(fulldata, nut_threshold = 1.5)

Arguments

Value

A named list of specs (names are nutrient codes), with title, ref_value and ref_label set.

Description

CG region a region code belongs to

Usage

own_cg_region(r, mapping = "mappingCG6.xlsx")

Arguments

Value

The code itself for CG regions, the CG region for countries, NA for anything else (WLD, DVD/DVG, countries outside CG regions).

Description

Bar chart, one panel per region.

Usage

plot_b4t_gap_closed(
  cmp,
  regions = c(b4t_regions, "WLD"),
  indicator_label = "target",
  baseline_id = crio_baseline_id,
  title = NULL,
  save = FALSE,
  outfile = NULL,
  plotdir = ".",
  width = 10,
  height = 6
)

Arguments

Value

A ggplot.

Description

(A) baseline gap; (B) gap under the B4T run; (C) change in value vs baseline; (D) share of the baseline gap closed.

Usage

plot_b4t_gap_maps(
  cmp,
  ctymap,
  b4t_scenario = "All CG regions",
  baseline_id = crio_baseline_id,
  direction = c("below", "above"),
  gap_labeller = scales::label_number(),
  delta_labeller = scales::label_number(),
  indicator_label = "indicator",
  target_year = 2050,
  save = FALSE,
  outfile = NULL,
  plotdir = ".",
  width = 10,
  height = 6
)

Arguments

Value

A ggarrange object.

Description

Fill = percent change vs baseline; tile text = absolute change (top) and percent change (bottom). Own-region effects sit on the diagonal.

Usage

plot_b4t_heatmap(
  cmp,
  regions = c(b4t_regions, "WLD"),
  direction = c("below", "above"),
  delta_labeller = scales::label_number(accuracy = 0.01, style_positive = "plus"),
  pct_labeller = scales::percent_format(accuracy = 0.01, style_positive = "plus"),
  baseline_id = crio_baseline_id,
  title = NULL,
  save = FALSE,
  outfile = NULL,
  plotdir = ".",
  width = 8,
  height = 5.5
)

Arguments

Value

A ggplot.

Description

Puts all indicators on one scale: gap = (value - target) / target for "below" targets and (target - value) / target for "above" targets, so a positive gap is short of target and zero or negative is target met. Lines are the baseline; markers are B4T in the panel's own CG region (open circle) and in all CG regions (cross). WLD and countries outside CG regions have no own-region run.

Usage

plot_gap_overview(
  specs,
  regions = c(b4t_regions, "WLD"),
  keep_above = NULL,
  target_year = 2050,
  baseline_id = crio_baseline_id,
  mapping = "mappingCG6.xlsx",
  title = "Distance to target",
  save = FALSE,
  outfile = NULL,
  plotdir = ".",
  width = 12,
  height = 8
)

Arguments

Value

A ggplot.

Description

(A) baseline value in baseline_year; (B) mean across scenarios in target_year; (C) baseline distance from target in target_year; (D) percent of scenarios hitting the target. Rest-of-world countries (row_countries) are greyed out.

Usage

plot_indicator_maps(
  df,
  summary_tbl,
  ctymap,
  value_col,
  baseline_id = crio_baseline_id,
  threshold = NULL,
  threshold_col = NULL,
  direction = c("below", "above"),
  indicator_label,
  value_labeller = scales::percent_format(accuracy = 1),
  target_year = 2050,
  baseline_year = 2025,
  save = FALSE,
  outfile = NULL,
  plotdir = ".",
  width = 10,
  height = 6
)

Arguments

Details

With a fixed threshold the value scales diverge at the target; with a dynamic threshold_col they use a simple gradient.

Value

A ggarrange object.

Description

Regions where some but not all scenarios hit the target.

Usage

plot_pct_hits(analysis, title = "% scenarios hitting target")

Arguments

Value

A ggplot.

Description

Plot the classification tree for one region

Usage

plot_region_tree(analysis, region_code, title = NULL)

Arguments

Value

Called for its side effect (base graphics plot).

Description

Heatmap of tree rules across regions

Usage

plot_rules_heatmap(
  analysis,
  outcome_filter = "hits",
  title = "Pathways by region",
  wrap_width = 30,
  base_size = 11,
  save = FALSE,
  outfile = NULL,
  plotdir = ".",
  width = 18,
  height = 14
)

Arguments

Value

A ggplot.

Description

Convenience wrapper: passes a spec from indicator_specs() or nutrient_quantity_specs() to plot_target_trajectories().

Usage

plot_spec_trajectories(spec, ...)

Arguments

Value

A ggplot.

Description

Line figure over time, one panel per region. Bands show the range across GCMs for each pathway (SSP x RCP, climate ensemble); the black line is the baseline; coloured lines are the B4T runs built on the baseline (same SSP, GCM and RCP); the dashed line is the target.

Usage

plot_target_trajectories(
  df,
  value_col,
  threshold = NULL,
  threshold_col = NULL,
  regions = c(b4t_regions, "WLD"),
  b4t_show = c("all", "own"),
  ref_value = NULL,
  ref_label = "reference",
  baseline_id = crio_baseline_id,
  mapping = "mappingCG6.xlsx",
  indicator_label,
  value_labeller = scales::label_number(),
  title = NULL,
  save = FALSE,
  outfile = NULL,
  plotdir = ".",
  width = 10,
  height = 7
)

Arguments

Details

Note: if all GCMs give the same value (e.g. hunger at the model's floor), a band has zero width and is not visible.

Value

A ggplot.

Description

Calls a GFSir reader (e.g. GFSir::group5()) on each GDX file, adds the scenario id and binds the results.

Usage

read_scenarios(files, fun, ...)

Arguments

Value

A data.table with all scenarios stacked.

Description

Re-applies add_scenario_cols() to every element of fulldata, e.g. to rename pathways or change the GCM classification without re-reading GDXs.

Usage

relabel_scenarios(fulldata, ...)

Arguments

Value

fulldata with updated scenario columns.

Description

Adult RNIs per capita per day, from Sherwin's GTAP paper. Units are in the nutrient code suffix (g, mg, mcg).

Usage

rni_lookup

Format

A tibble with columns Nutrients and RNI.

Description

Rest-of-world countries greyed out on maps

Usage

row_countries

Format

A character vector of IMPACT region codes.

Description

One file per element of fulldata. CSVs keep only food groups, their commodities and totals where a name column exists.

Usage

save_crio_data(fulldata, rdsdir = NULL, csvdir = NULL, suffix = "")

Arguments

Value

fulldata, invisibly.

Description

Scenario id from a GDX file path

Usage

scenario_id(gdx)

Arguments

Value

File name without folder and .gdx extension.

Description

Filters every element of fulldata that has scenario columns, so all later figures and tables use only the chosen runs. NULL keeps everything for that dimension.

Usage

select_scenarios(
  fulldata,
  ssp = NULL,
  rcp = NULL,
  gcm = NULL,
  b4t = NULL,
  ids = NULL
)

Arguments

Value

fulldata with filtered elements.

Description

Summarise tree rules by region and outcome

Usage

summarise_rules(rules_tbl)

Arguments

Value

One row per region with total cover, number of rules and the rules for "hits" and "misses".

Description

Built from the maps package, without Antarctica, dissolved into one outline. Cached after the first call.

Usage

world_outline()

Value

A SpatVector.