Semi-automated process for handling mismatches between Cliopatria and other Seshat data

[edwardchalstrey1]

Decide how to resolve the below:

• Some polities e.g. gb_british_emp_2 (ID = 99) have a mismatch between the years in the Cliopatria shape dataset and the linked Seshat polity
• The power transition data may not match between Seshat and Cliopatria "components"
• The "area" in Cliopatria (per shape) may not match the "Polity Territory" in Seshat

How not to do this:

• Resolve mismatches between Seshat polity year ranges and shape dataset year ranges edwardchalstrey1/seshat#28
• Ideas on how to resolve mismatches between Seshat polity year ranges and shape dataset year ranges edwardchalstrey1/seshat#98

[edwardchalstrey1]

TODO

• Discuss this with Jim & Pieter
• We also need to handle that fact that Seshat has "Power transitions" and Cliopatria also has "leaders" which are similar but may not be exact matches - see Roman Empire polities

[edwardchalstrey1]

I think this issue is for a much later date once Cliopatria is an established part of Seshat and people are suggesting edits

[edwardchalstrey1]

Update after Seshat workshop November 2024

• We should work with Seshat/Cliopatria RAs (Jenny and Jakob Z work on Cliopatria) to have an automated way to check mismatches
  • Matilda might already have some code which does this. See ✨ [Feature request] - Integrate code for generation of Time Series Dataset to Seshat API package seshat_api#26
  • Jim also mentioned he had some code for this
• If there ends up being some code that needs to be run to check this, we can probably add this to the Seshat docs