-- Name: VIEW _plots_20_tnrs_input_names; Type: COMMENT; Schema: public_validations; Owner: -

--

COMMENT ON VIEW _plots_20_tnrs_input_names IS '

The first check is done while checking up on Aaron''s quantitative validations

("comparing the results of the plot output queries run on VegBIEN to the data in

VegBank"). Several input-output queries include verbatim taxon names; if they

pass then Aaron is forming the verbatim name correctly. Even though it concerns

the verbatim name and not the resolved name, this first check is important. If

Aaron is messing up the verbatim name somehow, the resolved name will not be

correct.

';

--

-- Name: VIEW _plots_21_tnrs_output_names; Type: COMMENT; Schema: public_validations; Owner: -

--

COMMENT ON VIEW _plots_21_tnrs_output_names IS '

The second check should focus on determining if Aaron is reconstructing

morphospecies names correctly. The TNRS stores the morphospecies part of the

name (e.g., "sp.1") in the column "Unmatched_terms"; a bit of additional

manipulation is required to concatenated it with the resolved part of the name

to form the morphospecies string ("Miconia sp.1", "Carex ''fuzzy leaf''"). As I

found out only recently, Aaron does NOT renormalize the resolved names back to

the core database. He simple stores the verbatim TNRS output. Fully formed

morphospecies names are therefore not stored anywhere in the database, but are

reassembled on the fly when generating an analytical extract. Aaron''s rules for

doing this are embedded in a script somewhere, I have no idea where. For Brian

or Bob to assess if Aaron is re-concatenating morphospecies names correctly,

they will need to request analytical extracts for one or two of their plots, and

compare them to the names in their original databases. This second step

shouldn''t take long. If the resolved morphospecies name in Aaron''s analytical

extract looks correct, you''re done.

';

--