[Feature Request]: Multiple Treatment variables and multiple instrument variables in DoubleMLPLIV

[moses-1604]

Describe the feature you want to propose or implement

Presently, I know that Multiple Treatment variables and multiple instrument variables in DoubleMLPLIV are possible, but is it possible to assign an instrument for a specific treatment variable when there are multiple treatments and multiple instruments?

Propose a possible solution or implementation

No response

Did you consider alternatives to the proposed solution. If yes, please describe

No response

Comments, context or references

No response

[SvenKlaassen]

In the current implementation all instruments are treated equally.
For specific choices of instruments I would instead use separate PLIV models. This also allows more finegrained control on the used covariates in each model.
Is there a specific reason you would like to fit in a single PLIV model?