NEWS.md
mermaid_get_project_data() now uses CSV endpoints, which significantly speeds up getting data from projects with a large number of observations
This endpoint brings a few changes - empty values for text fields such as label, project_notes, site_notes, etc, are now NA instead of ""
Fields with multiple values (e.g. tags, management_parties, etc) are now separated using a comma (,') instead of a semi-colon (;`)
Standard deviations have been added for all columns that have averages in mermaid_get_project_data() - e.g. fishbelt sample events now has biomass_kgha_trophic_group_sd in addition to biomass_kgha_trophic_group_avg.
Benthic. PQT sample units and sample events data gain the column num_points_nonother, which counts the number of non-other points
Fishbelt observations gains the column assigned_transect_width_m
All data retrieved via mermaid_get_project_data() now contains the project_admins column
Sample units data from mermaid_get_project_data() no longer has the random id column
mermaid_get_project_data() functions by default, require settings covariates = TRUE
mermaid_get_project_sites() also gains covariate argument, which when set to TRUE returns covariates at the site levelmermaid_get_project_data()
mermaid_import_project_data()
mermaid_import_project_data() to take project, not project_id, consistent with other functionsmermaid_import_check_options() more manually.mermaid_get_summary_sites() as endpoint has been removed.NAs are written to '' to appear as blanks, not literal "NA" in mermaid_import_project_data().mermaid_get_summary_sampleevents() for getting aggregated metrics for all MERMAID surveys, by site, by date, and mermaid_get_summary_sites() for getting aggregated metrics for all MERMAID surveys, by site, for all dates.mermaid_import_get_template_and_options() to get a template and field options for importing any method into MERMAID.mermaid_import_check_options() for verifying that data being prepared for import matches the field options allowed.mermaid_import_field_options(), replaced by the above two functions (which are more feature rich and complete).mermaid_get_project_data() functions failed if andrello or beyer covariates were not present, and data for multiple projects is being selected (0.4.4 only fixed the single-project case).mermaid_get_project_data() functions failed if andrello or beyer covariates were not present.mermaid_get_project_data()).NA) values in mermaid_import_project_data() by automatically converting to an empty value, which is processed by the API as a NULL, before importing.Sample time values in mermaid_import_project_data() by automatically converting NA to "" before importing.mermaid_get_sites() and mermaid_get_managements() now require authorization.mermaid_import_project_data().mermaid_import_field_options() to check valid options for fields in import.mermaid_get_project_endpoint(), since the endpoints were removed from the underlying API.mermaid_get_reference() to include regions.mermaid_get_project_data() to automatically unpack any data frame columns. This affects the fishbelt, benthic PIT, and benthic LIT methods, for both sample units and sample events data. This is a breaking change, expected to affect existing code that uses the biomass_kgha_by_trophic_group, biomass_kgha_by_fish_family, and percent_cover_by_benthic_category columns in sample units, and their *_avg counterparts in sample events. Instead of these columns, results will now contain a column for subgroup - for example, instead of biomass_kgha_by_trophic_group there will be columns such as biomass_kgha_trophic_group_piscivore and biomass_kgha_trophic_group_planktivore.mermaid_get_reference() to provide enhanced reference data, returning actual values for e.g. fish family, sizes, groups, etc, instead of their internal IDs. The display column for the “fishspecies” reference has been renamed to species (#21).covariates are NULL).mermaid_get_project_data()).biomass_kgha_by_fish_family and biomass_kgha_by_fish_family_avg to fishbelt sample units and sample events, respectively (in mermaid_get_project_data()).httr::RETRY() after one failure if the status code indicates an unauthorized request; no need to retry in those cases.mermaid_get_project_data(), and additional fields available in Fish Belt and Benthic PIT endpoints.url argument from most external functions, since switching between prod and dev is more complicated than just changing the url - especially for authenticated endpoint calls. For now, switching between prod and dev requires installing from the main and dev branches, respectively. I will continue to explore making the token generation more robust for accessing both prod and dev, at which point the url argument will likely return!httr::RETRY() instead of httr::GET() to make functions more resilient to e.g. temporary API outages or timeoutsvctrs.HTTP_API_VERSION header that is not properly handled by the httr package (https://github.com/r-lib/httr/issues/590).count column to fishbelt observations (queried via mermaid_get_project_data(method = "fishbelt", data = "observations")).