NEWS.md

In bcgov/rems: Get Data from British Columbia's Environmental Monitoring System

rems (0.8.1)

• ask parameter now skips all verification (caching and updates) when set to FALSE (#67, #68)
• download_historic_data() now has more flexible options to update the historic database (principally to enable better updating behaviour in shinyrems). (#69, thanks @aylapear)
• removed capability to register/convert some units due to a bug in the C library underlying the {units} package (https://github.com/r-quantities/units/issues/301). (#70)
• added temporary warning message to connect_historic_db() to inform users of #79 where the timestamp is removed from the COLLECTION_START and COLLECTION_END columns. The issue will be resolved when next {duckdb} version is released.

rems 0.8.0

• Switched from using SQLite to DuckDB as a backend for the historic database. This allows fast creation of the database directly from the csv file downloaded from the B.C. Data Catalogue (#52).
• The data returned to R from the database is now returned with dates in Pacific Standard Time (PST; UTC -8), which matches the 2-year and 4-year datasets
• Removed ems_posix_numeric() function

rems 0.7.0

• Added new helper function standardize_mdl_units() to detect when MDL_UNIT and UNIT are different, and convert METHOD_DETECTION_LIMIT to the same unit as RESULT (and update the MDL_UNIT column accordingly). (https://github.com/bcgov/wqbc/issues/158, #57)
• Added new helper function lt_lake_req() that returns the requisition IDs of lakes in the B.C. long-term lakes monitoring program. To compliment this, filter_ems_data() now has a req_id argument that takes a vector of requisition ids (which can be supplied by lt_lake_req()). #60, @KarHarker
• lt_lake_sites() now only returns EMS IDs for active sites in the B.C. long-term lakes monitoring program. #60, @KarHarker
• Fixed a bug where the entire data cache was deleted when the historic data were updated. #62, #63

rems 0.6.1

Bug fixes

• Fixed a bug where read_historic_data() was looking for the date of the latest historic data on GitHub (#56)

rems 0.6.0

Historic data are now being updated daily in the BC Data Catalogue. This release allows the daily update of the sqlite historic dataset in rems, but it means that it must be recreated each time from the csv, which takes a while.

Breaking changes:

• When using attach_historic_data(), it is now necessary to call connect_historic_db() first, and finish with disconnect_historic_db(). For example:

con <- connect_historic_db()
hist_tbl <- attach_historic_data(con)
## Do your work with hist_tbl
disconnect_historic_db()

rems 0.5.2

• removed internal use of filter_ to avoid dplyr deprecation warnings. (#48)

rems 0.5.1

• Fixed bug where currency of the historic sqlite database was checked against the wrong date
• filter_ems_data can now take many more forms of date-like inputs for from_date and to_date (Date, POSIXct, POSIXlt, character, numeric; #42)
• Added some (mainly internal) functionality to support development of the shinyrems package (#43).

rems 0.5.0

• The historic sqlite database is now available as a direct download from a rems release on GitHub, which should be much faster and more reliable for the user.
• User can now specify/customize the location of the sqlite database via an option rems.historic.path. This is best set in your .Rprofile file as options(rems.historic.path = "path_to_file")

rems 0.4.2

• Added new MDL_UNIT column to denote the units of the minimum detection limit.

rems 0.4.1

• Added data dictionaries (lookup tables for parameters, sample classes, units, etc)
• Added dont_update argument to get_ems_data() and download_historic_data() to bypass the check to update data (#21).
• Added lt_lake_sites() function t get the EMS_IDs of all of the long-term lake monitoring sites (ac34dbd)
• Added check_only argument (default FALSE) to get_ems_data() to allow just checking the currency of a rems dataset (#35 @sebdalgarno)
• Added check_db argument (default TRUE) to read_historic_data() so that a user can skip checking the currency of the historic dataset (#35 @sebdalgarno)

rems 0.4.0

• Added indexes to several key columns in the 'historic' sqlite database. This makes loading the data slower, but makes queries and the read_historic_data function much faster.
• Added param_code argument to filter_ems_data and read_historic_data
• Added PERMIT, SAMPLE_CLASS, SAMPLE_DESCRIPTOR to default "wq" columns
• Renamed load_histori_data() to attach_historic_data()

rems 0.3.0

• You can now dowload the four most recent years of data, in addition to the most recent two years of data, as before.
  • This is a breaking change, as previous code that called get_ems_data(which = "current") will now throw an error. It now needs to be get_ems_data(which = "2yr") to achieve the same result as before, or it now can be get_ems_data(which = "4yr") to get four years of data.

rems 0.2.0

• Added new columns to data that were added in server files (#12)
• Added LOCATION_PURPOSE and SAMPLE_STATE to 'wq' columns (#13)
• Renamed get_update_date() to get_cache_date() along with some refactoring of code for getting/setting update dates (#11)
• Added ask parameter to get_ems_data() and download_historic_data() to optionally bypass download confirmation dialogue (#10)

rems 0.1.2

• Fixed bug where changes in file listings in DataBC catalogue would cause errors

rems 0.1.1

• Fixed bug where changes in file listings in DataBC catalogue would cause errors

rems 0.1.0

• Initial release

bcgov/rems documentation built on Oct. 14, 2023, 3:04 a.m.