| NEWS | R Documentation |
New option to exclude all records with any NA values for any of the
selected variables added to finbif_occurrence.
The warning/error message returned when an invalid taxon is
requested with finbif_occurrence has been made more detailed,
warning the user they may be using synonyms or other invalid taxon
names.
Fix applied for a print method error triggered by clashing class names.
Fixed issue with ordering by some variables.
New variables and filters (currently undocumented) have been added for the forthcoming Finnish atlas of breeding birds.
Skip rows argument added to finbif_occurrence_load so files
can be read in chunks.
For clarity, KKJ coordinate reference system is now refered to as YKJ.
When selecting all available variables for download file imports
using finbif_occurrence_load "extra" variables will no longer
be automatically added.
New administrative statuses added.
The number of (and interval between) api request retries can be configured (currently undocumented).
For some museum collections scientific names have been added for non-Finnish species.
New fields have been added to citable download imports.
New argument aggregate_counts (defaulting to TRUE)
added to finbif_occurrence to toggle the inclusion of
aggregation count fields on and off.
New bird association area, occurrence types and location tag metadata added.
New filters and variables (currently undocumented) added from upstream API.
Fix bug triggered when printing occurrence records where all records have no coordinate uncertainty.
Argument drop_na_facts has been changed to
drop_facts_na for function finbif_occurrence_load to
avoid problems due to partial matching of argument drop_na.
Data quality description has been appended to description field in
data.frame returned by finbif_collections().
Occurrence record filtering can now be done using open-ended date ranges.
New administrative statuses added.
Fix bug in importing of file downloads triggered by string quotation errors.
New package option, finbif_allow_query, with default value,
TRUE. If set to FALSE then remote API queries will trigger
an error meaning only cached requests will run if a cache is found.
Records can now be aggregated by document.
A client-side error will be triggered if an attempt is made to
filter by taxon and aggregate data by event with
finbif_occurrence.
Fix bug preventing loading of FinBIF archives that is triggered
when all variables are selected with select = "all".
Fix bug preventing finbif_occ objects from printing when one
or more columns contain missing data only.
Fixed a regression in the last release that prevented collection names being converted from their URIs.
New variables lat_euref and lon_euref available for
occurrence record downloads. These new variables represent the central
point of a bounding box encompassing a record's geographic coverage in
the EUREF (ETRS89/ETRS-TM35FIN) coordinate system.
Option now available to retain the TSV data file when loading FinBIF
occurrence data with finbif_occurrence_load from ZIP archive or
online reference.
Option now available to extract "facts" when using
finbif_occurrence_load to import records from a ZIP archive. The
user must supply a named list, where the names refer to one or more fact
types: "record", "event" and/or "document", and the list elements
are character vectors indicating which "facts" to extract.
"Lite" download files (occurrence record datasets of less than
10,000 records downloaded directly from laji.fi) can now be imported
using finbif_occurrence_load.
New filters (currently undocumented) ely_center_id and
region_id are available for finbif_occurrence.
New function, from_schema, to convert variable names from
FinBIF schema to other styles.
New computed variable date_time_ISO8601 (currently
undocumented) can be selected when creating finbif_occ objects.
New computed variable epsg (currently undocumented). If
selected, the column will have the EPSG code of the first column that
contains geographic data.
Can now select all available variables when importing a FinBIF
download request file using finbif_occurrence_load with
select = "all".
New option, drop_na, to drop columns that only contain
missing data added to finbif_occurrence and
finbif_occurrence_load.
User can now select columns multiple times and rename them on the fly by using a named character vector as the value of the 'select' argument.
The variable taxon_rank is now converted to a string on the
fly via taxon_rank_id.
Variables that are computed from their identifiers are now localised when multiple languages are available.
Fixed bug that prevented occurrence record counting when caching was not in use.
Fixed bug that triggered an error when selected a variable that had to be computed from its ID when using DWC style variable names.
Getting records from FinBIF can now be speed up (~1.5x) with
asynchronous processing. If the future package is available and
an asynchronous strategy (e.g., plan(multisession, workers = 2))
is selected then downloading records and processing them will occur
simultaneously in separate threads, often leading to a significant
speed up over sequential processing.
Occurrence records can now be requested using multiple sets of filters at the same time. If a list of filters (with unnamed elements) is supplied then a record request will be made for each set of filters and the results of all requests combined and all duplicate records removed.
Occurrence records can now be read directly from FinBIF download
request files. The function finbif_occurrence_load can load data
from a local file or remotely using a persistent identifier URI (e.g.,
http://tun.fi/HBF.49381).
Aggregation of records can now be performed at the recording "event" level as well as the "record" level.
New functions scientific_name and common_name are now
available to get scientific and common names for taxa (given a taxon
name or FinBIF ID code) or extract names from a finbif_taxa-
class object.
Add ability to extract complex record variables such as those associated with record media (currently undocumented).
Collection name can be selected as a variable for occurrence records. Previously it had to be manually translated from collection ID.
Variables restriction_reasons_conservation,
restriction_reason_embargo and restriction_reason_custom
are no longer in use upstream and can no longer be selected.
Occurrence record requests can now be made with aggregation,
aggregate = c("records", "species", "taxa"). This returns the
number of records, and/or species or taxa for each combination of the
selected variables instead of the occurrence records themselves.
Error is triggered when attempting to request less than one record.
New content on requesting aggregated occurrence records added to vignette: 2. Occurrence records from FinBIF.
Fixed two bugs in print method for FinBIF occurrence records that were triggered when attempting to print only one row of data.
Fixed bug triggering error when attempting to print occurrence record objects with zero rows.
Fixed bug triggered when trying to print occurrence record data with NA values.
Fixed bug causing incorrect subsetting of occurrence records when using logical vectors to subset rows.
Fixed bug that printed occurrence records with multi-element data incorrectly when there was a single element.
Fixed bug triggered by some system locales (#1).
Occurrence records can be ordered by the total number of records or
total number of Finnish records of the taxon (variables:
n_total_records & n_total_finnish_records, currently
undocumented).
New Finnish occurrence status, records_only, added.
New administrative statuses added to filters.
Filtering vignette updated to reflect changes to data quality filters.
All user facing functions with a finbif_ prefix can now also
be used with the alternative shorter prefix fb_.
Crop Wild Relative, CWR, added to admin statuses.
There are two new utility functions, to_dwc and
to_native for converting variable names to and from Darwin Core
style.
Users can select and order by variables in Darwin Core style when
using finbif_occurrence.
Some changes in variables and filters have flowed from upstream
changes to "api.laji.fi". The variables is_unidentifiable,
record_reliable, collection_reliability,
taxon_reliability, taxon_reliability_message and
taxon_reliability_source have been deprecated and replaced with
requires_verification, requires_identification,
record_reliability and record_quality. The filters
collection_reliability and taxon_reliability have been
deprecated and replaced with requires_verification,
collection_quality, record_reliability,
record_quality, and expert_verified.
Vernacular names are now localised. Users can select a language to use for taxon vernacular names. Missing names will fallback gracefully to other languages. A package-wide locale can be set and is by default set to the system locale (if not set or can't be determined it will default to English).
New vignettes on getting occurrence records, selecting and ordering variables, metadata and plotting have been added.
Caching defaults to in memory caching instead of relying on the temporary directory.
Front matter of vignettes is now visible when using the R help browser.
Retired "quiet" option for "on_check_fail" argument in function
finbif_occurrence().
Improved error messages when taxa fail checking in
finbif_occurrence().
Global option to set timezone, "finbif_tz", as default value for
"tzone" argument to finbif_occurrence(). Avoids having to set
system environment variable TZ or specify "tzone" every time
finbif_occurrence() is run.
Now when a record has no time information the start time is assumed to be midday. Previous behaviour was to assume start time was midnight, making errors potentially biased.
Package options are now documented in the package level man page.
News file is now accessible via R internal help system.
The number of default variables selected when accessing occurrence
records has been reduced to speed up downloads and improve the display
of finbif_occ objects.
The print method for occurrence record objects has been updated. It is now aware of console width and when truncating variable values is more considerate of the context.
Variables can now be "deselected" when using the select
argument to finbif_occurrence by prepending the variable name
with a "-".
After some failures, API requests are now automatically retried up to three times.
Ordering by descending variables did not work when ordering by both ascending and descending variables.
Fixed bug in handling of duplicates that could result in an infinite recursion.
Fixed bug that (when "on_check_fail" = "warn") all taxa failed checks 'finbif_occurrence()' would proceed as if no taxa had been selected.
Fixed bug in print method for finbif_occ objects that caused
error when trying to display a single column objects with a list-col
only.
Add capacity to request a random sample of FinBIF records.
Add more content to vignettes.
Initial release.