2013-12-19 conference call¶

Martha's notes¶

upcoming¶

• no calls during winter break, which includes 12/26 and 1/2
• the next conference call is Th. 1/9 at 9am PT/10am Tucson/12pm ET

availability¶

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

decisions¶

beta release¶

• not until January
  • deferred because we want all the critical issues fixed before the release

SALVIAS attribution¶

• contributors
  • need to include plot contributors in PlotMetadata.PrimOwnerID,plot_administrator
    • translate these usernames to full names, using Brad's new party_code_party, party lookup tables on nimoy in salvias_plots
  • project contributors (in projects.project_pi) don't need to be included separately because they are listed as plot contributors
  • attribution of contributors also happens by linking to the projects/plots in SALVIAS
• project names also need to be mapped to provide subprovider (institutional) attribution

GBIF validation¶

• missing stateProvince data is not an issue, because the dataset didn't provide this field
• missing families in TNRS results are an issue
  • the workaround for accepted names will be applied at the beginning of January
    • use Accepted_name_family when Name_matched_accepted_family is not available
      note that this workaround does not fix the name in the GBIF extract that led us to this problem (Ecdeiocolea monostachya), because that is a no-opinion name
  • no opinion names are not used in analyses, so the more-involved fix for these (which requires rescrubbing them) will be postponed
• the geoscrubbing isn't providing scrubbed placenames when there is no lat/long, but we're not sure which part of the pipeline this bug is in
  • need to fix this for the beta release

BIEN2 traits validation¶

• the extract needs to contain the data providers and locality information so these can be validated
  • we already have views that left-join all the tables together, so Brad doesn't need to write a query to do this
• we would like to get a refresh for this datasource from Cyrille, but this will take awhile because he has to hand-assemble it
  • the TRY data is combined with other data hand-entered from the literature, so we can't just submit a data request to TRY

FIA validation¶

• Brad needs the concatenated plot code, not just the numeric primary key in PLOT.CN
  • this is because the PLOT table actually contains one row per census, not one row per plot

to do for Aaron¶

for December¶

• fix the priority issues in VegBank/CVS/SALVIAS (from Brad's e-mail)
  • map VegBank/CVS/SALVIAS project names
  • map VegBank project contributors (CVS doesn't provide project contributors, according to Mike Lee)
  • provide project contributors in a separate validation view they have been concatenated together in the main view, which was easier to implement than an entirely new validation view

for January¶

• implement TNRS families workaround
• figure out why geoscrubbing isn't providing scrubbed placenames when there is no lat/long
• create a view that left-joins the traits to the observations, for use in the BIEN2 traits validation

long-term¶

• add TNRS and geoscrubbing tables to the ERD
  • TNRS.tnrs, TNRS views [redundant with TNRS.tnrs], geoscrub.geoscrub_output
• re-run TNRS on the no opinion names that are missing a family
• refresh traits data from Cyrille
  • this will have the same format as the current raw traits data on nimoy
    • the CSVs all have the same column headers, so they can just be appended