NEWS

fluxnet 0.6.0.9000

Added experimental functions flux_db_connect(), flux_db_build(), and flux_db_update() that ingest FLUXNET data into a local DuckDB database that can be queried with dplyr without reading data into memory.
Added a vignette to demonstrate working with the new DuckDB functions.

flux_qc() now works with hourly data by supplying a threshold of 0, 1, 2, or 3.
The max_gapfill argument to flux_qc() has been renamed to threshold
The output of flux_read() now includes a time_resolution column. This may be especially useful for hourly ("HR") and half-hourly ("HH") frequency data, which are combined when resolution = "h".

Added flux_citations() to generate site-level citations either as plain text or BibTex entries.
Fixed a bug where hourly data wasn't being extracted or read in along with half-hourly data (#73)

Fixes a bug introduced in 0.3.1 when more than one site id is provided to site_ids.

site_ids = "all" is deprecated in favor of site_ids = NULL in all functions with this argument.

Adds a new function, flux_install_shuttle() which accepts the name of a virtualenv to use and a version of fluxnet-shuttle to install. It pulls these from envvars, by default, making it easy to configure via a project-level .Renviron file. It can be run as a separate step, but this is optional as it will be called automatically the first time flux_listall() or flux_download() are run.
flux_download() no longer shows a progress bar as it downloads files.
Optionally, you can provide a list containing your Ameriflux user name, email, and intended use to flux_download() via the new user_info argument and flux_amf_credentials() helper function.
To get all available sites with flux_download(), use site_ids = NULL. site_ids = "all" is deprecated.
flux_download() now uses the fluxnet_shuttle Python library rather than httr2 to perform downloads.
Fixed a bug in flux_badm() that would cause it to error when some "BIF" files contained additional optional columns (#57)
The use_cache argument of flux_listall() has been removed.
flux_listall() gained a clean_cache argument to set the number of recent snapshot CSVs to keep. Defaults to 10.

Added flux_varinfo() and flux_badm() for reading and tidying "BIFVARINFO" and "BIF" files, respectively.
flux_extract() and flux_read() gain a network argument to specify a subset of networks (e.g. ICOS, TERN) to extract data from. Suggested by @lbell3141.
Added flux_qc(), a function to flag overly gapfilled rows of aggregated (not hourly) data.
Added flux_map_sites().
flux_download() now uses httr2 rather than curl for downloading files. This fixed a bug where downloads were failing when attempting to download large numbers of sites at once.
Added flux_map_sites()
flux_download() prints a warning when downloads fail.
Changed default cache age for flux_listall() to 1 day.
flux_download() now retries failed downloads once and overwrite = FALSE no longer skips downloading corrupted or partial downloads.
Added flux_read() for reading in FLUXNET data in a manifest.
Added flux_discover_files() for creating a file manifest.