Remove `k_endog` argument from statespace model constructors

[Dekermanjian]

Many of the State Space models have as inputs k_endog & endog_names and k_exog and exog_state_names as inputs that a user can optionally provide one or the other of the pair. While the option is nice to have the side effect is that there is a lot of additional validation code that needs to be added to check that at least one or the other is provided.

My proposal is to only provide endog_names and exog_state_names as required inputs to reduce the amount of validation code but also to make outputs of the model more specific to the user's data.

I discussed this with @jessegrabowski and we would like to hear others` thoughts on the matter.

[AlexAndorra]

Good idea Jonathan. I agree with the issue you raised, and the proposed solution El El mié, 1 oct 2025 a la(s) 7:20, Jonathan Dekermanjian < ***@***.***> escribió:

…

*Dekermanjian* created an issue (pymc-devs/pymc-extras#589) <#589> Many of the State Space models have as inputs k_endog & endog_names and k_exog and exog_state_names as inputs that a user can optionally provide one or the other of the pair. While the option is nice to have the side effect is that there is a lot of additional validation code that needs to be added to check that at least one or the other is provided. My proposal is to only provide endog_names and exog_state_names as required inputs to reduce the amount of validation code but also to make outputs of the model more specific to the user's data. I discussed this with @jessegrabowski <https://github.com/jessegrabowski> and we would like to hear others` thoughts on the matter. — Reply to this email directly, view it on GitHub <#589>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHIJMTCBPBXAXGF7VICCYXL3VO2ILAVCNFSM6AAAAACH72GNU6VHI2DSMVQWIX3LMV43ASLTON2WKOZTGQ3TGMJSGUZDEOA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>