Skip to contents

From obs_bio joined to taxon, with lifeStage from life_stage.csv and the D8 denominator as organismQuantity:

Usage

dwc_occurrence(
  con,
  dataset_key,
  life_stage = NULL,
  measurement_type = NULL,
  absences = c("none", "sample_root"),
  max_absences = 5e+06
)

Arguments

con

a DBI connection to the release

dataset_key

the dataset

life_stage

the registry from read_life_stage_registry(), or NULL

measurement_type

the registry from read_measurement_type(), or NULL

absences

"none" or "sample_root"

max_absences

refuse to derive more absences than this

Value

A data frame of Occurrence rows, all-NA columns dropped.

Details

Darwin Corecore
occurrenceIDmd5 of (sample_key, taxon_key, life_stage, measurement_type, depth_bin, ordinal) — stable across releases (obs_id is not)
eventIDobs_bio.sample_key, or the root's sample_key for a derived absence
scientificName / taxonID / taxonRank / lineagetaxon
scientificNameIDthe WoRMS LSID of taxon.worms_id; empty when the taxon has none
lifeStagelife_stage.csv dwc_lifeStage, else the verbatim value where the registry gives it a life_stage_parent; a value the registry records as not a life stage goes to occurrenceRemarks
organismQuantity / organismQuantityTypedensity_per_10m2, else density_per_1000m3, else value + its registry units
occurrenceStatus"present" where value > 0, "absent" where value = 0

The absence rule. See dwc_absence_rule(). absences = "none" (the default) emits only rows the release holds. absences = "sample_root" additionally emits one absent row for every (surveyed root sample x observed taxon/stage) pair with no positive row — the sample_root minus positives rule. It is correct only for a dataset whose protocol sorts every sample for its whole vocabulary, so it is never the default, and max_absences (5,000,000) stops a large vocabulary from turning a survey into a hundred million assertions nobody made.