Skip to contents

Plan the list of COG fetches

Source: R/collect_tern_data.R
dot-build_work_items.Rd

A work item is one COG-shaped request: a tuple of (dataset, date_idx, variant, depth). Each item maps to exactly one read_tern() + terra::extract() round-trip and produces one or more columns in the output table.

Usage

.build_work_items(
  datasets,
  dates,
  depth,
  stat,
  smips_collection,
  asc_collection,
  aet_collection,
  soildiv_collection,
  phenology_collection,
  canopy_collection
)

Arguments

datasets

Normalised alias vector.

dates

Resolved Date vector.

depth

SLGA depth selector.

stat

SLGA stat selector.

smips_collection

SMIPS collection selector.

asc_collection

ASC collection selector.

aet_collection

AET collection selector.

soildiv_collection

SOILDIV collection selector.

phenology_collection

PHENOLOGY collection selector.

canopy_collection

CANOPY collection selector.

Value

A list of work-item lists.

Details

For time-series datasets (SMIPS, AET) we emit one work item per (date, variant); for PHENOLOGY datasets, which are annual over two growing seasons, we emit one work item per (year, variant, season) for each date; for SLGA datasets we emit one work item per (depth, variant) combination. For temporally-static datasets, the item values are replicated across the date axis (date_idx = NA_integer_).

date_idx is therefore an integer vector of positions into dates, and NA_integer_ marks an item whose single value applies to every date.