Possible categorization error: dataset try

[beatrizmilz]

Hello!

OTN is a great project, thank you all for it.

This issue aims to document a possible error in the "resolved" categorization.

While using the dataset, Thiago @thiago-goncalves-souza and I noticed a possible categorization error on the try dataset (https://opentraits.org/datasets/try).

If we filter OTN to get only rows that are from the try dataset AND Animalia Kingdom (resolveKingdomName == "Animalia"), we get more than 5k rows.

# download data from
# https://github.com/open-traits-network/otn-taxon-trait-summary/blob/main/traits.csv.gz
otn_raw <-
  readr::read_csv("traits.csv")

otn_dataset_try <- otn_raw |>
  # filter only the animal kingdom
  dplyr::filter(resolveKingdomName == "Animalia") |>
  dplyr::filter(datasetId == "https://opentraits.org/datasets/try")


dplyr::glimpse(otn_dataset_try)
# Rows: 5,311
# Columns: 31
# $ taxonIdVerbatim        <chr> "1669", "1669", "1669", "1669", "1669", "1…
# $ scientificNameVerbatim <chr> "Agathis philippinensis", "Agathis philipp…
# $ resolvedTaxonId        <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ resolvedTaxonName      <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ parentTaxonId          <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ family                 <chr> "Araucariaceae", "Araucariaceae", "Araucar…
# $ phylum                 <chr> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ traitIdVerbatim        <dbl> 37, 3400, 759, 98, 3401, 43, 22, 17, 4, 38…
# $ traitNameVerbatim      <chr> "Leaf phenology type", "Plant growth form …
# $ bucketId               <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ bucketName             <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ counts                 <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ datasetId              <chr> "https://opentraits.org/datasets/try", "ht…
# $ numberOfRecords        <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 1, 1, 3, …
# $ curator                <chr> "https://opentraits.org/members/brian-s-ma…
# $ accessDate             <date> 2022-08-19, 2022-08-19, 2022-08-19, 2022-…
# $ comment                <chr> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ relationName           <chr> "HAS_ACCEPTED_NAME", "HAS_ACCEPTED_NAME", …
# $ resolvedExternalId     <chr> "COL:6635V", "COL:6635V", "COL:6635V", "CO…
# $ resolvedName           <chr> "Agathis philippinensis", "Agathis philipp…
# $ resolvedRank           <chr> "species", "species", "species", "species"…
# $ resolvedCommonNames    <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA…
# $ resolvedPath           <chr> "Biota | Animalia | Arthropoda | Insecta |…
# $ resolvedPathIds        <chr> "COL:5T6MX | COL:N | COL:RT | COL:H6 | COL…
# $ resolvedPathNames      <chr> "unranked | kingdom | phylum | class | ord…
# $ resolvedExternalUrl    <chr> "https://www.catalogueoflife.org/data/taxo…
# $ resolveKingdomName     <chr> "Animalia", "Animalia", "Animalia", "Anima…
# $ resolvedPhylumName     <chr> "Arthropoda", "Arthropoda", "Arthropoda", …
# $ resolvedFamilyName     <chr> "Braconidae", "Braconidae", "Braconidae", …
# $ providedTraitName      <chr> "Leaf phenology type", "Plant growth form …
# $ resolvedTraitName      <chr> "Phenology", "Morphology", "UNCATEGORIZED_…

But some of the traits seems like they are from plants:

otn_dataset_try |>
  dplyr::count(datasetId,
               resolveKingdomName,
               providedTraitName,
               sort = TRUE) |> 
  head() 

datasetId	resolveKingdomName	providedTraitName	n
https://opentraits.org/datasets/try	Animalia	Plant growth form	482
https://opentraits.org/datasets/try	Animalia	Leaf type	257
https://opentraits.org/datasets/try	Animalia	Leaf compoundness	255
https://opentraits.org/datasets/try	Animalia	Plant woodiness	255
https://opentraits.org/datasets/try	Animalia	Leaf phenology type	178
https://opentraits.org/datasets/try	Animalia	Leaf area (in case of compound leaves: leaflet	161

Here are some of the most frequent categories that appear in resolvedPhylumName/resolvedName from this query:

otn_dataset_try |>
  dplyr::count(datasetId,
               resolveKingdomName,
               resolvedPhylumName,
               resolvedName,
               sort = TRUE) |> 
  head()

datasetId	resolveKingdomName	resolvedPhylumName	resolvedName	n
https://opentraits.org/datasets/try	Animalia	Mollusca	Ficus	162
https://opentraits.org/datasets/try	Animalia	Chordata	Salix	118
https://opentraits.org/datasets/try	Animalia	Arthropoda	Eugenia	117
https://opentraits.org/datasets/try	Animalia	Arthropoda	Inga	117
https://opentraits.org/datasets/try	Animalia	Arthropoda	Viola	94
https://opentraits.org/datasets/try	Animalia	Chordata	Phyllanthus	88