2012-08-17 conference call¶

To do¶

1. Switch all datasources to VegCSV's custom sort order
2. Complete loading of all remaining BIEN2 data as soon as possible
   1. Make table indicating which data has been loaded
   2. Load raw CTFS data (not VegX!)
3. Begin building analytical database with help from Brad (once core VegBIEN database is completely loaded)
4. Have analytical database, containing all data sources (plots and specimens) plus completed validations, ready by next BIEN meeting at end of November

VegCSV changes¶

1. Add verbatim prefix to taxon determination terms for original determination
2. Remove computer prefix from taxon determination terms
3. Add namespaces to terms indicating provenance provenance is instead indicated next to the term name
   • Requires expanding the mapping algorithm to match terms with and without a namespace

VegBIEN schema changes¶

1. locationDetermination: Add verbatimCoordinates, verbatimLatitude, verbatimLongitude text fields to store original coordinates

Other topics¶

How does VegCSV handle data constraints?¶

• Dependency relationships (hierarchies) are modeled in the import order of the CSVs
• Constraints are provided by the VegBIEN schema as well as by the data provider's own SQL DDL

Analytical database¶

• Based on *bien_web.observation, Brad's improvement to BIEN2's viewFullOccurrence*
• Discussion tabled until next meeting

E-mail from Brad on 2012-8-17:

"The scripts that created table bien_web.observation are in bien_shared/bien_web on nimoy.

Additional analytical tables in bien_web contain alternate summaries of bien2 data, optimized for different analyses. These are bien_results, bien_results_nwemb, bien_species_results, bien_species_results_nwemb. The scripts that created them are also in directory bien_web, each in subdirectories bearing the same name as the table they create."