The TRY file is huge and unnecessarily long, which makes it difficult to work with. The resulting SQLite database is much smaller on disk, and can be read much faster thanks to lazy evaluation.

try2sqlite(try_files, sqlite_file = "try.sqlite")

Arguments

try_files

Character vector of file names containing TRY data. Multiple files are combined with `data.table::rbindlist`.

sqlite_file

Target SQLite database file name, as character.

Details

The resulting TRY SQLite database contains the following tables: - `values` -- The actual TRY data. Links to all other tables through ID columns. - `traits` -- Description of trait and data names. Links to `values` through `DataID`. Similar to BETY `variables` table. - `datasets` -- Description of datasets and references/citations. Links to `values` through `DatasetID` and `ReferenceID`. - `species` -- Species. Links to `values` through `AccSpeciesID`.