2014-01-23 conference call¶

Martha's notes¶

upcoming¶

• the next conference call is next week at the usual time (Th. 9am PT/10am Tucson/12pm ET)

availability¶

• please add your availability for spring 2014 to the *spreadsheet*:

to do for Mike Lee¶

aggregating validations¶

• write input-side aggregating queries for VegBank and CVS
  • these should produce output in the same format as the VegBIEN aggregating queries for plots

to do for Brad¶

aggregating validations¶

• write input-side aggregating queries for plots datasources other than VegBank/CVS: FIA, SALVIAS, Madidi, TEAM, CTFS, NVS

decisions¶

aggregating validations¶

• for specimens only, OK for the queries to use the renamed VegCore column names (Brad)
  • the plots queries must be run with the datasource's original column names
• for specimens only, OK for the aggregating validations to validate just the normalization aspect of the import process (rather than also the mappings) (Brad)
• OK for input rows to have been deleted before the aggregating validations are run (e.g. due to frameshift errors or private data) (Brad)
• OK to make schema changes to make writing the queries easier (Brad, Martha) (Martha: unless schema change would take too long)

to do for Aaron¶

SALVIAS aggregating validations¶

1. produce sample result set of running the input-side queries
2. make schema changes to simplify the SQL for nested plots
   1. modify the locationevent__pull_forward_from_parent() trigger function to populate the project_id from the parent plot event
   2. allow locationevents (and therefore taxonoccurrences) to be reached directly from top-level plots, to allow aggregating things by top-level plot rather than just subplot
3. write the blank queries
   • these were the ones that were difficult to create (Brad)
4. fix the queries that have issues: #12, 18
   • indicated by all-caps notes in the query comment
5. change the taxonomic queries to use the verbatim name in taxonlabel.taxonomicname instead of concatenating the ranks together manually
6. translate MySQL input queries to Postgres so they can run on the staging tables

plots aggregating validations¶

1. ensure that column unrenaming is working not using this method anymore
2. automate validations as part of import process

specimens aggregating validations¶

1. rename columns in Brad's NY queries to VegCore so that they can be used for all specimens datasources
2. create the variation for each specimens datasource
   • this involves replacing non-existent columns with NULL