Add a proper "repeat measurements id"

[ehwenk]

There are now two groups who will be using traits.build who require a repeat measurements counter to keep track of individual measurements on curve data. The current method of creating replicate counters in custom_R_code, then mapping this in as a temporal context works but is not a good solution if it has to be added to every dataset - not least because it isn't a context and errors can easily be introduced if users don't group by the correct collection of columns.

Instead, it should be automated. Thoughts:

• repeat_measurements_id (or some other name) will be added to austraits$traits
• repeat_measurements_id = FALSE will be added to metadata_create_template - because otherwise I think this will mask when there should be actual contexts added; in other words, default is to not have repeat_measurement_id's enabled
• if repeat_measurements_id = TRUE, we'll add something to process.R that groups by all the current "need to pivot" columns, then adds a counter if nrow > 1 for any grouping. Columns are:

dataset_id, trait_name, observation_id, source_id, taxon_name, entity_type, life_stage, basis_of_record, value_type,
         population_id, individual_id, temporal_id, method_id, entity_context_id, original_name

[ehwenk]

A few things here, some duplicated from above notes:

• I started a branch, but I think it is best if NO default changes occur to metadata_create_template; can delete those. It just won't appear in the metadata file unless needed.
• People could add repeat_measurements_id: TRUE at either the dataset or trait level - at the trait level if the entire dataset is response curves of some sort.
• As a default, within process.R, repeat_measurements_id = NA, with row number counters added if flipped to true.
• Need to add repeat_measurements_id to schema (I think true on my branch)
• Requires that in the pivot test, the ncol goes up 1 again, and repeat_measurements_id added to list of vars to select and pivot on

[yangsophieee]

Resolved with this commit