SEED Mapping

Mapping in SEED

Mapping refers to the process of renaming columns in the input table to match standard or expected names such as (and primarily) those specified by the Building Energy Data Exchange Specification, or BEDES. There are currently two different ways this mapping can be performed:

1. For data that is known to be from the Portfolio Manager program, a "fixed" mapping is performed where the input columns are looked up against a static specification (e.g., seed/data/pm-mapping.json). This is done in the common/mapping.py module.
2. For data from any other source, a "fuzzy" mapping is used. The program uses some version of the edit distance to determine which known column name approximates the one in the input data, and gives up if nothing passes a hard-coded threshold.

Both of these branches are initiated from the REST API method get_column_mapping_suggestions in the Django view module seed/views/main.py.

This method also retrieves the previous mapping, if any, for this building and organization. The user-defined column edits in this mapping will override anything determined from the algorithms in 1. and 2. above.