The goal of gbifexplorer is provide some tools to explore the different
types of gfbif dataset classes:
Occurrence, Sampling-event or Checklist. gbifexplorer allows to
explore the taxonomic, temporal and/or spatial coverages of the gbif
datasets.
You can install the development version of gbifexplorer from GitHub with:
# install.packages("devtools")
devtools::install_github("ajpelu/gbifexplorer")The gbifexplorer package provides several functions for exploring and
analyzing occurrence data from the Global Biodiversity Information
Facility (GBIF). Here is an example of the usage of three key functions:
cov_temporal, taxonomic_cov, and report_taxonomy.
We will use a dataset contains information about the phenology of flora
in Mediterranean high-mountains meadows in the Sierra Nevada region
(Spain). This dataset is deposited in
GBIF and also published as Data
Paper. It also is included
as example data in the gbifexplorer pkg.
The gbifexplorer::cov_temporal() function calculates the temporal
coverage of a dataset based on a specified date variable provided by the
user. It determines the minimum and maximum dates from the date variable
and returns them in a data frame.
library(gbifexplorer)
data("borreguiles")
# Calculate temporal coverage
temporal_coverage <- cov_temporal(borreguiles, date_var = "eventDate",
date_format = "%Y-%m-%d")
#> The temporal coverage of the dataset is 1988-05-18 to 2013-10-17.It returns a data frame (temporal_coverage) that contains the minimum
and maximum dates from the dataset. This function also prints the
temporal coverage as text and could be used to document the metadata of
the dataset.
The gbifexplorer::taxonomic_cov() function generates the taxonomic
coverage of a dataframe. It calculates the record numbers and relative
frequencies of each taxonomic category specified, allowing to compute
the taxonomic coverage of an occurence dataset. The function returns a
named list of tibbles summarizing the taxonomic coverage for each
category.
# Calculate taxonomic coverage for scientificName and genus
d <- taxonomic_cov(borreguiles, category = c("class", "order"))
d
#> $class
#> # A tibble: 3 × 3
#> class n freq
#> <chr> <int> <dbl>
#> 1 Magnoliopsida 6057 55.1
#> 2 Liliopsida 4882 44.4
#> 3 Psilotopsida 63 0.573
#>
#> $order
#> # A tibble: 19 × 3
#> order n freq
#> <chr> <int> <dbl>
#> 1 Poales 4869 44.3
#> 2 Lamiales 1378 12.5
#> 3 Fabales 1190 10.8
#> 4 Asterales 1032 9.38
#> 5 Gentianales 996 9.05
#> 6 Ranunculales 460 4.18
#> 7 Caryophyllales 425 3.86
#> 8 Celastrales 209 1.90
#> 9 Malpighiales 200 1.82
#> 10 Ericales 103 0.936
#> 11 Ophioglossales 63 0.573
#> 12 Apiales 31 0.282
#> 13 Saxifragales 17 0.155
#> 14 Boraginales 9 0.0818
#> 15 Liliales 8 0.0727
#> 16 Asparagales 5 0.0454
#> 17 Brassicales 3 0.0273
#> 18 Rosales 3 0.0273
#> 19 Myrtales 1 0.00909The result is a list contains tibbles summarizing the taxonomic coverage for each taxonomic category.
Then, it is possible to generate a Report for different taxa categories
by using the gbifexplorer::report_taxonomy() function. It generates a
summary report of the taxonomy of a specified taxa rank, and provides
information about the most represented taxa based on the frequency
information of the taxa rank. The default is the top 5 most represented
taxa but is could be change using the argument top
For instance, if you are interested in the 5 most frequent order:
report_taxonomy(d$order, top = 5)
#> There are 19 order included in the dataset. The 5 order most represented in the dataset are: Poales (44.26 %), Lamiales (12.52 %), Fabales (10.82 %), Asterales (9.38 %) and Gentianales (9.05 %).The user might also be interested in calculate the taxonomic coverage
for all taxonomic categories present in an occurrence dataset. For this,
we also used the purrr package
library(purrr)
all_taxa <- taxonomic_cov(borreguiles, category = "all")
all_taxa |>
purrr::map(~report_taxonomy(., top=10)) |>
purrr::list_transpose()
#> [[1]]
#> There are 94 scientificName included in the dataset. The 10 scientificName most represented in the dataset are: Nardus stricta L. (9.35 %), Carex nigra (L.) Reichard (6.67 %), Euphrasia willkommii Freyn (6.47 %), Lotus corniculatus L. subsp. glacialis (Boiss.) Valdés (5.49 %), Scorzoneroides (5.32 %), Eleocharis quinqueflora (Hartmann) O.Schwarz (4.54 %), Festuca iberica (Hack.) Patzke (4.46 %), Carex nevadensis Boiss. & Reut. (4.38 %), Gentiana boryi Boiss. (3.7 %) and Plantago nivalis Jord. (2.92 %).
#>
#> All records belong to the kingdom Plantae
#> There are 2 phylum included in the dataset. The 2 phylum most represented in the dataset are: Magnoliophyta (99.43 %) and Pteridophyta (0.57 %).
#>
#> There are 3 class included in the dataset. The 3 class most represented in the dataset are: Magnoliopsida (55.05 %), Liliopsida (44.37 %) and Psilotopsida (0.57 %).
#>
#> There are 19 order included in the dataset. The 10 order most represented in the dataset are: Poales (44.26 %), Lamiales (12.52 %), Fabales (10.82 %), Asterales (9.38 %), Gentianales (9.05 %), Ranunculales (4.18 %), Caryophyllales (3.86 %), Celastrales (1.9 %), Malpighiales (1.82 %) and Ericales (0.94 %).
#>
#> There are 28 family included in the dataset. The 10 family most represented in the dataset are: Cyperaceae (21.12 %), Poaceae (19.52 %), Fabaceae (10.82 %), Asteraceae (9.01 %), Gentianaceae (8.82 %), Scrophulariaceae (7.22 %), Ranunculaceae (4.18 %), Caryophyllaceae (3.75 %), Juncaceae (3.61 %) and Plantaginaceae (2.92 %).
#>
#> There are 52 genus included in the dataset. The 10 genus most represented in the dataset are: Carex (16.58 %), Nardus (9.35 %), Scorzoneroides (9 %), Gentiana (8.8 %), Euphrasia (6.47 %), Lotus (5.49 %), Trifolium (5.33 %), Festuca (4.77 %), Eleocharis (4.54 %) and Agrostis (4.44 %).We also have developed a shiny app to allow the user to generate the taxonomic coverage of an occurence data. To run the app:
library("gbifexplorer")
gbifexplorer::taxo_reportApp()