Which datasets actually fed the scored surface?

A release's dataset table registers every dataset that was ingested, which is not the same as every dataset the scores were computed from. v8 ingests NOAA SEFSC (gm) and NCCOS (nc) density models into dist/, but their density units (#/km²) are not yet mapped onto the 0,100 suitability scale, so they are deliberately excluded from the merge — and therefore contribute nothing to any score. Listing them beside AquaMaps as though they did overstates the inputs.

Usage

dataset_is_scored(ds_key, taxon_model_ds_key, merged_key = "ms_merge")

Arguments

ds_key

character vector of dataset keys to classify

taxon_model_ds_key

ds_key column of that release's taxon_model (the taxon ↔ contributing-model edges); NULL when the release does not publish the relation, in which case every registered dataset is assumed scored — the honest answer when the release gives no way to tell

merged_key

the merged dataset's key

Value

logical vector parallel to ds_key

Details

The test is introspected, never asserted: a dataset is scored iff it contributed at least one model to a merged taxon, i.e. its ds_key appears in taxon_model. The merged dataset itself (ms_merge) is scored by construction — it is the surface scoring reads.

Examples

dataset_is_scored(c("ms_merge", "am", "gm"), c("am", "am", "bl"))  # TRUE TRUE FALSE
#> [1]  TRUE  TRUE FALSE