Remove the default configurations and attributes

[mgdenno]

I think we should consider removing the default configurations and attributes from the template. I think it just makes it messy. Might be more useful if we knew that each value in say configurations was actually used.

[mgdenno]

@samlamont what say you?

[samlamont]

Sounds good to me, I agree it's messy and potentially confusing for the user. How about the units and variables tables as well? That way everything in the dataset is actually being used by a particular evaluation. Maybe we could automatically load domain values if they are members of hard-coded enums otherwise require the user to add manually?

I was also thinking about creating another level of sub-directories in the dataset, in an attempt to provide more clarity for the user. Something like:

• timeseries
  • primary
  • secondary
  • joined
• domains
  • units
  • variables
  • configurations
  • attributes
• spatial/locations
  • locations
  • locations crosswalks

Maybe this just adds unnecessary complexity?

[mgdenno]

I think units and variables, esp. units are a but more complicated. Since we ultimately want to move towards automatic unit conversions i think we need consistency there...