README.md

forestNETNarch

This package was originally called forestNETN and was designed to connect with the original Access-based NETN forest backend database. It is now being archived and forestNETN is being updated to work with the new SQL-based backend.

The first step to using this package is to either execute the importData() or importCSV() function, which loads the most commonly used database tables into the global environment and names the tables accordingly.

The importData() function uses an odbc driver for MSAccess to pull tables directly from the NETN forest database. This requires an MSAccess odbc driver be installed on the machine, and the version of R (32 vs 64 bit) must match the version of Access. In the case of the NETN forest database, this should be 32-bit R. This function can either use the master NETN forest database named "NETNFVM" as a SystemDSN, or a database path can be specified. This may only work on Windows-based operating systems, and requires odbc and DBI packages.

The importCSV() function imports comma separated values files that were exported from the NETN Forest Database. As long as the user has these exported .csv tables, this function will import and load all of the commonly used tables into the global environment and name the tables the same as the importData() function.

The joinLocEvent(), joinTreeData(), joinRegenData(), joinQuadData(), and joinCWDData() source the tables from the global environment, and create common views of the data. These functions also allow the user to filter by common factors, such as park, years, exotic or native species, etc.



KateMMiller/forestNETNarch documentation built on Dec. 18, 2021, 2:41 a.m.